What does “methodology” mean?

“methodology” means the whole set of ways a study is carried out, and the reasons for them. Pronounced /ˌmɛθəˈdɑlədʒi/. It is formal. Close synonyms: approach, procedure, research plan. For example: The methodology section explains how participants were chosen and which tests were run on them.

What does “research design” mean?

“research design” means the overall plan for how a study will answer its question. Pronounced /ˈrisɜrtʃ dɪˈzaɪn/. It is formal. Close synonyms: study design, research plan, setup. For example: A cross-sectional research design cannot tell us whether attitudes changed over the students' four years.

What does “rationale” mean?

“rationale” means the set of reasons for doing something in a particular way. Pronounced /ˌræʃəˈnæl/. It is formal. Close synonyms: reasoning, justification, grounds. For example: The authors give a clear rationale for excluding students who had already taken the course.

What does “protocol” mean?

“protocol” means a written plan of steps that everyone in a study must follow. Pronounced /ˈproʊtəkɔl/. It is formal. Close synonyms: set of rules, standard procedure, plan. For example: All three laboratories followed the same protocol, so the results can be compared directly.

What does “qualitative” mean?

“qualitative” means based on words, meanings and descriptions rather than numbers. Pronounced /ˈkwɑləˌteɪtɪv/. It is formal. Close synonyms: descriptive, non-numerical, interpretive. For example: The qualitative interviews revealed reasons for dropping out that the survey numbers had hidden.

What does “quantitative” mean?

“quantitative” means based on numbers and amounts that can be counted or measured. Pronounced /ˈkwɑntəˌteɪtɪv/. It is formal. Close synonyms: numerical, statistical, measurable. For example: Quantitative results showed a small drop in errors, though the interviews told a richer story.

← Back to Public Decks · English Decks · English vocabulary test

English C1 — Academic & Research Official

The English of papers, studies and argument at C1, defined in English. Words for describing method, evidence, causation and limitation, each with an English definition, IPA, synonyms and a usage note on the hedging and precision academic writing demands.

CEFR C1 — Advanced

At this level you can use language flexibly, understand implicit meaning, work in target language. Vocabulary: ~6,000 words (cumulative).

Words in this deck

What does “methodology” mean?: “methodology” means the whole set of ways a study is carried out, and the reasons for them. Pronounced /ˌmɛθəˈdɑlədʒi/. It is formal. Close synonyms: approach, procedure, research plan. For example: The methodology section explains how participants were chosen and which tests were run on them.
What does “research design” mean?: “research design” means the overall plan for how a study will answer its question. Pronounced /ˈrisɜrtʃ dɪˈzaɪn/. It is formal. Close synonyms: study design, research plan, setup. For example: A cross-sectional research design cannot tell us whether attitudes changed over the students' four years.
What does “rationale” mean?: “rationale” means the set of reasons for doing something in a particular way. Pronounced /ˌræʃəˈnæl/. It is formal. Close synonyms: reasoning, justification, grounds. For example: The authors give a clear rationale for excluding students who had already taken the course.
What does “protocol” mean?: “protocol” means a written plan of steps that everyone in a study must follow. Pronounced /ˈproʊtəkɔl/. It is formal. Close synonyms: set of rules, standard procedure, plan. For example: All three laboratories followed the same protocol, so the results can be compared directly.
What does “qualitative” mean?: “qualitative” means based on words, meanings and descriptions rather than numbers. Pronounced /ˈkwɑləˌteɪtɪv/. It is formal. Close synonyms: descriptive, non-numerical, interpretive. For example: The qualitative interviews revealed reasons for dropping out that the survey numbers had hidden.
What does “quantitative” mean?: “quantitative” means based on numbers and amounts that can be counted or measured. Pronounced /ˈkwɑntəˌteɪtɪv/. It is formal. Close synonyms: numerical, statistical, measurable. For example: Quantitative results showed a small drop in errors, though the interviews told a richer story.

English · definitions in English 201 words ~21 days at 10 words/day 0 downloads C1

No ratings No ratings by admin@wordsonrepeat.com Sep 29, 2026

Try studying — no account needed

A real study session, right now. No signup.

Report issue with this deck

Word List — click grammar notes to expand

Word	Translation	Example
methodologyformal /ˌmɛθəˈdɑlədʒi/ the whole set of ways a study is carried out, and the reasons for them Synonyms: approach, procedure, research plan	the set of methods used	The methodology section explains how participants were chosen and which tests were run on them.
Methodology is the reasoning behind the methods; one step in a study is a method, not a methodology. Writers say 'the methodology used here' or 'a qualitative methodology'. Using it as a grander word for a single technique is the error editors flag most.
research designformal /ˈrisɜrtʃ dɪˈzaɪn/ the overall plan for how a study will answer its question Synonyms: study design, research plan, setup	the plan of a study	A cross-sectional research design cannot tell us whether attitudes changed over the students' four years.
It names the shape of the study, who is compared with whom and when, not the tools used to collect information. After the first mention it is usually shortened to 'the design'. Countable: 'a within-subjects research design'.
rationaleformal /ˌræʃəˈnæl/ the set of reasons for doing something in a particular way Synonyms: reasoning, justification, grounds	the reasons behind a choice	The authors give a clear rationale for excluding students who had already taken the course.
Use 'the rationale for X' or 'the rationale behind X', not 'of X'. Countable but almost always singular. It answers the question 'why this choice', so a rationale can be weak or convincing but never true or false.
protocolformal /ˈproʊtəkɔl/ a written plan of steps that everyone in a study must follow Synonyms: set of rules, standard procedure, plan	a fixed set of steps	All three laboratories followed the same protocol, so the results can be compared directly.
In research you 'follow', 'deviate from' or 'depart from' a protocol. The study protocol is written before collection begins, which is why any later change has to be reported. A separate sense covers formal official behavior.
qualitativeformal /ˈkwɑləˌteɪtɪv/ based on words, meanings and descriptions rather than numbers Synonyms: descriptive, non-numerical, interpretive	based on words, not numbers	The qualitative interviews revealed reasons for dropping out that the survey numbers had hidden.
It pairs with quantitative, and the two differ in kind of material, not in value. Frames: 'qualitative data', 'a qualitative study', 'qualitative work'. It never means 'of high quality', which is the mistake to avoid.
quantitativeformal /ˈkwɑntəˌteɪtɪv/ based on numbers and amounts that can be counted or measured Synonyms: numerical, statistical, measurable	based on numbers	Quantitative results showed a small drop in errors, though the interviews told a richer story.
Frames: 'quantitative data', 'a quantitative approach', 'quantitative methods'. It describes the material, not the quality of the work, so a 'quantitative question' is simply one answered with numbers. In casual American speech the middle consonant is often swallowed.
mixed methodsformal /ˌmɪkst ˈmɛθədz/ using both number-based and word-based ways of studying something Synonyms: combined approach, multi-method design	numbers plus interviews	The team took a mixed methods approach, pairing a national survey with thirty long interviews.
Always plural, and used before a noun with no hyphen in most American journals: 'a mixed methods study'. The term implies the two kinds of material are brought together in the argument, not merely collected side by side.
to operationalizeformal /ˌɑpəˈreɪʃənəˌlaɪz/ decide exactly how an idea will be measured in a study Synonyms: define in measurable terms, specify, pin down	turn into something measurable	The authors operationalize well-being as a score on a six-item questionnaire.
The frame is 'operationalize X as Y': the broad idea first, the measure second. British journals spell it 'operationalise'. The noun is 'operationalization', among the longest words a method section will ask a reader to take in.
constructformal /ˈkɑnstrʌkt/ an idea such as trust or worry that cannot be seen and must be measured indirectly Synonyms: concept, idea, underlying quality	an idea being measured	Motivation is a construct, so any single test can capture only part of it.
The noun stresses the first syllable, the verb 'to construct' the second. Countable, and usually tied to a measure: 'the construct of resilience', 'construct validity'. Keep it apart from the everyday sense of building something.
parameterformal /pəˈræmətər/ a fixed value or limit that shapes how a model or study works Synonyms: fixed value, setting, boundary	a fixed value in a model	The model has three parameters, each estimated from the historical rainfall record.
In statistics a parameter describes the whole group while a statistic describes the cases measured. In looser writing 'within the parameters of' means 'within the limits of', a use many editors still resist. The plural is far commoner than the singular.
eligibilityformal /ˌɛlədʒəˈbɪləti/ the state of meeting the rules for being included in something Synonyms: qualification, entitlement, fitness to join	the right to be included	Eligibility was limited to patients who had received no earlier treatment for the condition.
In a method section it pairs with exclusion: the eligible cases are those the rules let in, and the next sentence usually says who was then left out. The frames are 'eligibility for' and 'eligibility to do' something. The adjective is eligible, the negative ineligible.
cohortformal /ˈkoʊhɔrt/ a group of people studied together over time because they share a starting point Synonyms: group, set of participants, generation	a group followed over time	The 1958 birth cohort has been followed for more than sixty years.
What holds a cohort together is a shared moment in time, not shared treatment. Frames: 'a cohort of 4,000 nurses', 'a cohort study'. In casual use it can mean a companion, often with disapproval, so the research sense needs a clear setting.
longitudinalformal /ˌlɑndʒəˈtudənəl/ following the same people or cases over a long period Synonyms: long-term, over-time, follow-up	following the same cases over time	Only longitudinal work can show whether early reading ability predicts later income.
Frames: 'a longitudinal study', 'longitudinal data'. Its opposite is cross-sectional, a single snapshot. Length alone does not make a study longitudinal; the same cases must be measured more than once.
cross-sectionalformal /ˌkrɔs ˈsɛkʃənəl/ looking at many cases at a single point in time Synonyms: one-time, snapshot, single-point	taken at one point in time	Because the survey was cross-sectional, it cannot tell us which change came first.
A near-fixed limitation sentence follows it: a cross-sectional design can show that two things go together but not which came first. Contrast longitudinal, where the same cases are measured again. Hyphenated in American journals.
baselineneutral /ˈbeɪslaɪn/ the first measurement, kept for comparison with later ones Synonyms: starting point, first measurement, reference point	the first measurement	Reading scores at baseline were similar in both groups, which makes the later gap easier to read.
The frame 'at baseline' takes no article, and the word also works as a modifier in 'baseline scores'. Later values are compared 'with' or 'to' baseline. It marks a point in time, not a low value.
pilot studyformal /ˈpaɪlət ˌstʌdi/ a small early study done to test the plan before the full one Synonyms: trial study, small-scale test, trial run	a small trial run	A pilot study with twelve volunteers showed that the questionnaire took too long to finish.
Its job is to check that the method works, so its numbers are not usually offered as results. You 'run' or 'carry out' a pilot study, and the verb exists too: researchers pilot an instrument before using it.
interventionformal /ˌɪntərˈvɛnʃən/ something new that a study introduces to see what effect it has Synonyms: program, treatment, measure	the new thing introduced	Schools that received the intervention reported fewer absences in the following term.
In research it need not be medical: a lesson plan or a text message can be one. The frames are 'deliver', 'receive' and 'evaluate an intervention'. Keep it apart from the everyday sense of stepping into a quarrel.
treatment (in a study)formal /ˈtritmənt/ the condition a group is given in a study so it can be compared Synonyms: condition, program, intervention	what a group receives	Participants in the treatment group wrote for ten minutes a day; the others did nothing.
In experimental writing the word covers only what a group receives, which may be a lesson plan or a text message rather than anything medical. Frames: 'the treatment group', 'treatment effect', 'assigned to treatment'. Countable when several are compared.
conditionformal /kənˈdɪʃən/ one of the situations a study sets up so that groups can be compared Synonyms: experimental setting, group, treatment	one version of a task	In the noisy condition, participants remembered about two fewer words than in silence.
Normally named rather than left bare: 'the control condition', 'across conditions', 'in the delayed condition'. It is not the medical sense of a lasting illness, nor the 'on condition that' of agreements.
placeboformal /pləˈsiboʊ/ something with no real effect, given so that a comparison is fair Synonyms: dummy treatment, inactive pill	a treatment with no effect	Half the volunteers took the drug and half took a placebo made of sugar.
A placebo has to look like the real treatment, or the comparison it supports breaks down. 'Placebo-controlled' describes the design, and the 'placebo effect' is improvement caused by belief alone. The plural is placebos.
double-blindformal /ˌdʌbəl ˈblaɪnd/ arranged so that neither the people studied nor the staff know who got what Synonyms: masked, blinded	neither side knows the groups	The trial was double-blind, so the doctors scoring the outcomes could not guess the groups.
Single-blind hides the groups from participants only, so say which you mean. The related verb is 'to blind' and the noun 'blinding'. Hyphenated both before a noun and after 'was': 'a double-blind design', 'the study was double-blind'.
to randomizeformal /ˈrændəˌmaɪz/ put people into groups purely by chance, so the groups start out alike Synonyms: assign by chance, allocate by lot	assign purely by chance	Patients were randomized to either the new drug or standard care.
The frame is 'randomize X to Y', nearly always in the passive. Chance here is deliberate and controlled rather than careless. British journals write 'randomise'; the noun is randomization.
to stratifyformal /ˈstrætəˌfaɪ/ divide a group into smaller groups that share a feature before choosing from each Synonyms: divide into layers, group by category	divide into layers first	The sampling was stratified by age so that older workers were not left out.
The aim is to guarantee that every layer appears rather than trusting chance, which is why it is used when a small group would otherwise be missed. Usually passive and followed by 'by': 'stratified by region'. Noun: stratification.
to assign (to a group)neutral /əˈsaɪn/ place a participant into one of the groups being compared Synonyms: allocate, place in a condition	place into a study group	Volunteers were assigned to one of three conditions by a computer.
Random assignment is what the word promises in a method section: a computer or a coin decides, not the researcher. The frame is 'assign X to Y', and the noun is assignment. Assigning by convenience and then calling it random is the slip reviewers look for.
populationformal /ˌpɑpjəˈleɪʃən/ the whole group a study aims to describe, not only those measured Synonyms: whole group, target group, universe	the whole group studied	The clinics studied were unusual, so the results may not hold for the wider population.
In research it means everyone the conclusion is meant to cover, and it may be very small; the everyday sense of a country's people is a different use. Frames: 'the target population', 'generalize to the population'.
subsetformal /ˈsʌbsɛt/ a smaller group taken from inside a larger one Synonyms: smaller group, portion, part	a smaller part of a group	Only a subset of the responses, those from full-time staff, was analyzed here.
Because the smaller group comes from within the larger, a subset cannot contain anything the whole does not, and it is always 'a subset of' something named. Give it a singular verb: 'a subset of the responses was analyzed'.
samplingformal /ˈsæmplɪŋ/ the way cases are chosen from a larger group for study Synonyms: selection, case selection	how cases are chosen	Snowball sampling produced a group of interviewees who mostly knew one another.
A 'sampling error' is the natural gap between the cases chosen and the whole group, not a mistake anyone made. The word names the process, not the cases: 'random sampling', 'sampling took place in March'. Uncountable.
respondentformal /rɪˈspɑndənt/ a person who answers the questions in a survey Synonyms: survey participant, interviewee, answerer	someone who answers a survey	Two thirds of respondents said they had never heard of the program.
Standard in survey writing, where 'participant' suits experiments better. Frames: 'respondents reported', 'of the 400 respondents'. The word also has a legal sense, the person answering a claim in court.
questionnaireneutral /ˌkwɛstʃəˈnɛr/ a set of written questions used to collect answers from many people Synonyms: survey form, set of questions, measure	a written set of questions	The questionnaire took eight minutes and asked about sleep, diet and exercise.
The questionnaire is the paper or screen; the survey is the whole study. In formal writing you 'administer', 'complete' or 'return' one rather than 'make' one. Note the double n in the spelling.
survey (as a method)neutral /ˈsɜrveɪ/ a study that collects answers from many people in order to describe a group Synonyms: poll, questionnaire study	a study asking many people	A survey of 2,000 households found that a third had changed heating supplier.
As this noun the stress falls on the first syllable. Patterns: 'a survey of' a group, 'in a survey', 'survey data'. Because it records what people say rather than what they do, writers pair it with the hedge 'self-reported'.
fieldworkformal /ˈfildwɜrk/ research done in the real place being studied rather than in a laboratory Synonyms: field research, on-site research	research done on site	Her fieldwork in three small fishing villages lasted a total of eleven months.
It implies the researcher was present where the events happen, and 'in the field' points to the same setting. You 'carry out' fieldwork or work 'during fieldwork'. Written as one word, with no plural.
to transcribeformal /trænˈskraɪb/ write down exactly what was said in a recording Synonyms: write out, type up, put into writing	write out a recording	All forty interviews were transcribed in full before the coding began.
It covers speech turned into writing and never means translate, which is the error worth guarding against. Usually passive in method sections: 'were transcribed word for word'. The act is transcription, the result a transcript.
informed consentformal /ɪnˌfɔrmd kənˈsɛnt/ agreement given by someone who has been told what a study involves Synonyms: permission, voluntary agreement	agreement after being told	Written informed consent was obtained from every participant before testing began.
The word 'informed' carries the weight, since it claims the person understood the risks and was not merely willing. Consent is 'obtained from' or 'given by' people, never 'made', and there is no plural.
anonymityformal /ˌænəˈnɪməti/ the state of not being named or able to be identified Synonyms: namelessness, not being identified	not being named	Participants were promised anonymity, so no school is named in the report.
Anonymity means even the researcher cannot link an answer to a person, while confidentiality means they can but will not say, so promise only what the design allows. You 'guarantee', 'preserve' or 'protect' it. Adjective: anonymous.
confidentialityformal /ˌkɑnfəˌdɛnʃiˈæləti/ the promise to keep information private and not pass it on Synonyms: privacy, secrecy of records	keeping information private	Confidentiality was maintained by storing the recordings on a locked drive.
Use it where the researcher knows who said what but will not reveal it, which is the usual case in interview work. Frames: 'maintain', 'ensure' and 'breach confidentiality'. Adjective: confidential.
attritionformal /əˈtrɪʃən/ the gradual loss of participants from a study before it ends Synonyms: dropout, wastage, loss of cases	people leaving a study	Attrition was high in the third year, when a quarter of the families moved away.
It threatens a comparison because those who leave are often unlike those who stay, so a high rate has to be discussed and not just reported. Frames: 'the attrition rate', 'attrition was 18 percent'. A wider sense is slow wearing away.
reliabilityformal /rɪˌlaɪəˈbɪləti/ how consistently a test gives the same result in the same conditions Synonyms: consistency, dependability, stability	consistency of a measure	The scale showed high reliability, with almost identical scores two weeks apart.
In measurement it means consistency, not trustworthiness: a reliable test can be consistently wrong. Read it together with validity, which asks whether the right thing is being measured at all. No plural.
validityformal /vəˈlɪdəti/ how well a measure really captures what it is meant to capture Synonyms: soundness, fit, correctness	measuring the right thing	The authors question the validity of using test scores as a measure of teaching quality.
Always ask valid for what purpose, since nothing is valid in general, and the word is usually narrowed: 'internal validity', 'external validity', 'construct validity'. A study can be reliable without being valid, but not the reverse.
reproducibilityformal /ˌriprəˌdusəˈbɪləti/ the chance that others following the same steps reach the same result Synonyms: repeatability, verifiability	getting the same result again	Sharing the code and the raw numbers improves the reproducibility of the analysis.
Many American journals keep reproducibility for getting the same answer from the same material, and replication for a fresh study with new material, so the two are not interchangeable. Frames: 'concerns about reproducibility', 'the reproducibility crisis'.
to replicateformal /ˈrɛpləkeɪt/ repeat a study in the same way to see whether the result holds Synonyms: repeat, reproduce, do again	repeat a study	Two later teams failed to replicate the effect with much larger groups.
You replicate a study, a finding or an effect, and 'failed to replicate' is the standard way to report a negative outcome. It means repeat, not copy out. The noun is replication, and a study doing it is 'a replication'.
to standardizeformal /ˈstændərˌdaɪz/ make something follow the same fixed form every time Synonyms: make consistent, fix, regularize	make uniform	Instructions were standardized so that every group heard exactly the same wording.
In statistics it takes a narrower sense, rescaling values so that the average becomes zero, which is why 'a standardized score' is a technical term and not just a tidy one. Often passive: 'the wording was standardized'. British spelling: standardise.
to calibrateformal /ˈkæləbreɪt/ adjust an instrument so that its readings are correct Synonyms: adjust, fine-tune, set correctly	adjust for correct readings	The sensors were calibrated each morning against a known reference weight.
'Calibrate X against Y' names the standard used for the adjustment. Beyond instruments, writers calibrate a model to observed values, and figuratively calibrate how strong a claim should be. Noun: calibration.
apparatusformal /ˌæpəˈrætəs/ the set of equipment used to carry out a study or test Synonyms: equipment, setup, instruments	equipment for a study	The apparatus consisted of a screen, a response box and two sets of headphones.
Normally uncountable and taking a singular verb even though it names many parts: 'the apparatus was placed'. The plural 'apparatuses' exists but is rare. Apparatus is also the traditional heading in a written-up experiment.
specimenformal /ˈspɛsəmən/ a single item taken from something so that it can be examined Synonyms: sample, individual item, piece	a single sample item	Each blood specimen was frozen within an hour of collection.
It suits physical things taken for examination, such as tissue, rock or a pressed plant; survey answers and numbers are never specimens. Plural specimens. 'A fine specimen of' in speech is usually humorous.
instrumentformal /ˈɪnstrəmənt/ a tool used to collect measurements or answers in a study Synonyms: measure, tool, measuring device	a tool for measuring	The instrument used here is a twelve-item scale developed for older readers.
In social research an instrument is usually a questionnaire or test rather than a machine. Frames: 'develop', 'validate' or 'administer an instrument'. The adjective 'instrumental' means something else, helpful in bringing about.
to administerformal /ədˈmɪnəstər/ give a test, questionnaire or treatment to someone Synonyms: give, carry out, deliver	give a test to someone	The memory task was administered by a helper who did not know the group assignments.
'Administer X to Y', very often in the passive. It is the expected verb for giving a test, a drug or a survey and sounds more careful than 'give'. A separate sense is to run an organization.
to obtainformal /əbˈteɪn/ get something, often after effort or a formal request Synonyms: get, acquire, secure	get, often formally	Permission to obtain the hospital records was granted by two separate committees.
Do not use it for reaching a place or a level, which is the error it invites; you obtain permission, records or a result. Extremely common in the passive: 'consent was obtained', 'the records were obtained from'. Formal.
to screenneutral /skrin/ check people or items to decide which ones can be included Synonyms: check, test, filter	check for suitability	Applicants were screened for hearing loss before the listening test began.
'Screen X for Y' names what you are looking for, and 'screen out' removes those who do not qualify. It suggests a quick first check rather than a full examination. The noun is screening, as in 'a screening questionnaire'.
sample sizeneutral /ˈsæmpəl ˌsaɪz/ the number of people or cases included in a study Synonyms: number of participants, count of cases	how many cases studied	A sample size of thirty is small for a question with this much variation between people.
Written n in tables and reported as 'n = 240'. A larger sample narrows the range around an estimate but cannot repair a biased one, which is the confusion worth avoiding. Working it out in advance is called a power calculation.
datumformal /ˈdeɪtəm/ one single measurement or fact from a larger set Synonyms: single value, single measurement, reading	a single piece of information	Every datum in the table was checked twice against the original paper record.
The plural is data, /ˈdeɪtə/, and in American academic writing that plural now takes either a plural verb ('the data suggest') or a singular one ('the data is'); journals in the sciences prefer the plural. The singular datum is rare outside statistics.
datasetformal /ˈdeɪtəˌsɛt/ a collection of measurements stored together for study Synonyms: data file, collection of records	an organized collection of data	The dataset covers every recorded earthquake above magnitude four since 1950.
Most American journals now write it as one word, though 'data set' survives, and it takes a singular verb: 'the dataset contains', never 'contain'. You draw on, merge, clean or release a dataset.
metricformal /ˈmɛtrɪk/ a chosen way of measuring something so that cases can be compared Synonyms: measure, yardstick, index	a chosen measure	Citation counts are a poor metric for the influence of a young researcher.
A metric is chosen by the writer, so it can be criticized as poor or misleading, which a plain measurement cannot. Frames: 'a single metric', 'across three metrics'. The separate adjective 'metric' refers to meters and grams.
order of magnitudeformal /ˈɔrdər əv ˈmæɡnəˌtud/ a difference of about ten times in size Synonyms: roughly ten times, a factor of ten	a factor of about ten	The new estimate is an order of magnitude larger than the one published in 1998.
Used with a comparative: 'an order of magnitude larger', 'two orders of magnitude smaller'. The related phrase 'on the order of' means approximately, as in 'on the order of 10,000 cases'. It is a deliberately rough comparison.
to quantifyformal /ˈkwɑntəˌfaɪ/ measure or express something as an amount Synonyms: measure, put a number on, count	express as a number	The cost to families is real but hard to quantify from the records available.
Common in hedged admissions: 'difficult to quantify', 'we cannot quantify'. The noun is quantification, and the adjective 'quantifiable' often appears as the negative 'not readily quantifiable'. Its object is normally a noun phrase, though an indirect question also occurs: 'hard to quantify how much of the fall is real'.
to tabulateformal /ˈtæbjəleɪt/ arrange numbers or facts in rows and columns Synonyms: put in a table, arrange, list	arrange in a table	Responses were tabulated by age group and are tabulated again in the appendix.
It means only arranging in a table, so it makes no claim about what the numbers mean. Usually passive in results sections: 'results are tabulated below'. The noun is tabulation, and 'cross-tabulate' compares two categories at once.
to plotneutral /plɑt/ mark values on a graph so a pattern can be seen Synonyms: graph, chart, mark out	draw on a graph	When life expectancy is plotted against income, the curve flattens after about 20,000 dollars.
The frame is 'plot X against Y', with the outcome first and the explaining measure second. Very often passive: 'the values were plotted'. Quite separate senses are a secret plan and the story of a novel.
to aggregateformal /ˈæɡrəɡeɪt/ combine separate amounts into one total or group Synonyms: combine, pool, add together	add up into one figure	Scores were aggregated across the four tasks to give a single overall measure.
'Aggregate X across Y' is the frame, as in 'aggregated across regions'. The verb ends in a full vowel, /-eɪt/, while the noun and adjective end in /-ət/. Aggregating hides differences inside the group, which is why careful writers report the spread as well as the total.
meanformal /min/ the average found by adding values and dividing by how many there are Synonyms: average, arithmetic average	the arithmetic average	The mean waiting time was eleven days, though most patients waited far less.
As a noun it is countable: 'the group means differed'. Writing prefers 'mean' to 'average' where the median is also reported, since 'average' is vague between the two. A few extreme values pull the mean but barely move the median.
medianformal /ˈmidiən/ the middle value when all values are put in order Synonyms: middle value, midpoint	the middle value	The median income is a better guide here, since a few very high earners distort the mean.
Used as a noun and before other nouns: 'the median', 'median age', 'median household income'. Half the cases lie on each side of it, which is why it resists extreme values. In American English it also names the strip between highway lanes.
standard deviationformal /ˈstændərd ˌdiviˈeɪʃən/ a number showing how far values usually lie from the average Synonyms: spread, typical variation	typical spread around the mean	Both groups scored about 60, but the standard deviation was twice as large in the younger group.
A large standard deviation means the cases are spread out, not that the measurement was poor. It is abbreviated SD and usually reported in parentheses after a mean: 'M = 60.2 (SD = 8.1)'.
varianceformal /ˈvɛriəns/ a measure of how much values differ from one another Synonyms: spread, variation, scatter	how much values scatter	The model explains about a third of the variance in reading scores.
The frame to learn is 'explain the variance in X', usually given as a percentage. It is uncountable within a single analysis, but statisticians do count it when comparing groups, as in 'equal variances assumed'. The everyday sense of a disagreement survives in 'at variance with', meaning in conflict with.
to skewneutral /skju/ make a set of values lean to one side, or make a result unfair Synonyms: distort, bias, tilt	pull out of balance	A handful of very long stays skews the average, so the median is reported instead.
In statistics 'skewed' describes a lopsided spread and is purely descriptive; outside statistics it usually suggests something has gone wrong, so a reader may take it as criticism. Often passive: 'the sample was skewed toward older workers'. Noun: skew or skewness.
outlierformal /ˈaʊtˌlaɪər/ a value far away from all the others in a set Synonyms: extreme value, exception, odd case	an unusually extreme case	Two outliers were removed before the analysis, and both decisions are reported in a footnote.
An outlier may be an error or a real extreme case, and deciding which is a judgment, not a calculation. You identify, exclude, retain or keep outliers, and honest writing says which was done and why.
intervalneutral /ˈɪntərvəl/ the space or time between two points or values Synonyms: gap, space, period	a gap between two points	Measurements were taken at five-minute intervals throughout the four-hour session.
The fixed frame is 'at intervals of X' or 'at X intervals'. In measurement theory 'interval data' means the gaps between values are equal, which is what allows averaging. British English also uses it for a break in a performance.
confidence intervalformal /ˈkɑnfədəns ˈɪntərvəl/ a range that probably contains the true value Synonyms: likely range, range of uncertainty	a likely range for a value	The effect was small, and the confidence interval crossed zero, so caution is needed.
A wide interval signals uncertainty, and one that includes zero means no effect cannot be ruled out, which is the whole reason it is reported. Abbreviated CI, usually with a percentage: 'a 95 percent confidence interval'.
significant (in statistics)formal /sɪɡˈnɪfəkənt/ unlikely to have happened by chance alone Synonyms: reliable, not due to chance	unlikely to be chance	The difference was statistically significant, though it amounted to only two points.
In everyday English the word means important; in a results section it means only that chance is an unlikely explanation. A significant difference can be tiny, so careful writers add 'statistically' and switch to 'substantial' when they mean large.
significanceformal /sɪɡˈnɪfəkəns/ the degree to which a result is unlikely to be chance Synonyms: reliability, weight, importance	how far chance is ruled out	The authors report significance at the five percent level for two of the four measures.
The wider sense of importance also exists, so 'the significance of this finding' can mean either and the context has to settle it. Frames: 'reach significance', 'at the five percent level', 'tests of significance'.
to correlateformal /ˈkɔrəleɪt/ change together, so that one rises as the other rises or falls Synonyms: go together, be linked, vary together	go together, rise together	Hours of sleep correlate with test performance, but neither clearly causes the other.
'Correlate with', never 'correlate to', and the two things named can be swapped without changing the claim. That symmetry is why correlation cannot show which came first, the caution behind the familiar warning about cause.
coefficientformal /ˌkoʊəˈfɪʃənt/ a number in a formula or model showing how much one thing changes with another Synonyms: multiplier, weight, factor	a number in a formula	The coefficient on schooling is positive but falls by half once income is added.
A coefficient is read for its sign as well as its size, since a negative one reverses the direction of the claim. The preposition is 'on': 'the coefficient on age'. 'Correlation coefficient' names the best-known kind.
regressionformal /rɪˈɡrɛʃən/ a method that estimates how one measure changes with others Synonyms: statistical model, fitted model	fitting a predictive line	A simple regression suggests that each extra year of study adds about eight percent to earnings.
Things are 'regressed on' the explaining measures, and you 'run' a regression: 'in the second regression'. A separate sense, going back to an earlier and worse state, appears in 'regression to childhood'.
to control forformal /kənˈtroʊl fɔr/ allow for the effect of something else so it cannot explain the result Synonyms: adjust for, allow for, hold constant	hold other causes steady	Once the authors control for parental income, the advantage of private schooling almost disappears.
Always 'control for' plus the thing set aside, and the phrase 'controlling for age and sex' is a fixed part of results writing. Statistical control is not the same as a control group; it happens in the analysis, not the design.
covariateformal /koʊˈvɛriət/ something measured alongside the main interest because it may affect the result Synonyms: background measure, control variable	a background measure adjusted for	Age and prior scores were entered as covariates in every model reported here.
A covariate is not what the study is about; it is included so that it cannot be blamed for the pattern found. The fixed frames are 'enter X as a covariate' and 'adjust for covariates'.
biasneutral /ˈbaɪəs/ something in a study that pushes results in one direction Synonyms: slant, distortion, one-sidedness	a built-in slant	Asking only current members introduces an obvious bias toward satisfied customers.
In research bias is a property of a design, not an accusation about a person's honesty, and it is countable when named: 'selection bias', 'recall bias'. You introduce, reduce or correct for bias. 'Biased toward' is the usual adjective frame.
precisionformal /prɪˈsɪʒən/ how exact and repeatable a measurement is Synonyms: exactness, fineness, consistency	how finely something is measured	The newer device offers greater precision, reporting weight to the nearest gram.
Precision is about the fineness and repeatability of a reading, accuracy about whether it is right: a scale that always reads two grams heavy is precise but not accurate. 'With precision' is the usual adverbial frame.
accuracyneutral /ˈækjərəsi/ how close a measurement is to the real value Synonyms: correctness, closeness to truth	closeness to the truth	The accuracy of self-reported height was checked against measurements taken in the clinic.
Accuracy needs a true value to compare against, which is why it is checked against a second, better source. Precision is a different quality, about repeatability. Frames: 'the accuracy of X', 'with reasonable accuracy'.
to approximateformal /əˈprɑksəmeɪt/ be close to something without being exactly the same Synonyms: come close to, resemble, be near	be nearly the same as	The simulated pattern approximates the observed one closely enough for planning purposes.
The verb takes a direct object with no preposition: 'approximates the true value', not 'approximates to'. The verb ends /-eɪt/ while the adjective 'approximate' ends /-ət/. It admits a gap without saying how large it is.
false positiveneutral /ˌfɔls ˈpɑzətɪv/ a result that reports an effect which is not really there Synonyms: a spurious finding, wrong detection	a wrong yes	With twenty tests run on the same data, at least one false positive is likely.
Its partner is a false negative, a real effect the test misses, and saying which one you fear tells the reader how you set your threshold. Testing many things at once multiplies false positives, which is the argument behind correcting for multiple comparisons.
variabilityformal /ˌvɛriəˈbɪləti/ the degree to which things of the same kind differ from each other Synonyms: variation, inconsistency	how much values differ	There was wide variability between schools, even among those with similar intakes.
Between-group and within-group variability are the two a reader wants separated, since a difference in means says little when cases inside each group differ more. Wide, high and considerable are the usual adjectives. Variation is the plainer everyday word.
effect sizeformal /ɪˈfɛkt ˌsaɪz/ a number showing how large a difference is, not just whether it exists Synonyms: size of the difference, magnitude of an effect	how big the difference is	The effect size was small, which matters more for policy than the low p-value does.
Set it against significance, which only says an effect is unlikely to be chance: a significant result can have an effect size too small to care about. Reported as 'a small effect size' or with a named measure. Journals increasingly require one beside every test.
robustformal /roʊˈbʌst/ still holding true when the checks or conditions are changed Synonyms: sturdy, stable, holding up	holding up under checks	The result is robust to leaving out the two largest cities.
The academic frame is 'robust to X', where X is a change that might have broken the finding. A 'robustness check' repeats the analysis another way. The everyday sense of physically strong or vigorous is separate.
consistent withformal /kənˈsɪstənt wɪð/ fitting or agreeing with something without proving it Synonyms: in agreement with, in line with, matching	agreeing with, not proving	These patterns are consistent with the idea that attention narrows under stress.
One of the most useful hedges in academic English: it says the result fits an explanation while leaving room for others. 'Consistent with' takes a noun or a clause with 'the idea that'. Its negative is 'inconsistent with', not 'unconsistent'.
percentileformal /pərˈsɛnˌtaɪl/ the point below which a given percent of the cases fall Synonyms: rank position, hundredth place	a rank out of a hundred	A child at the ninetieth percentile reads better than ninety percent of the class.
It reports position, not amount, so a move from the fiftieth to the sixtieth percentile says nothing about how many points were gained. 'At the ninetieth percentile' is the usual frame. The median is simply the fiftieth percentile.
proxyformal /ˈprɑksi/ something measured as a stand-in for what you really want to measure Synonyms: stand-in, substitute measure, indirect measure	a stand-in measure	Years of schooling is used here as a proxy for skill, which is hard to observe directly.
The frame is 'a proxy for X', and it is an admission as much as a method: the real quantity was out of reach. Countable, plural proxies. A separate sense is a person who votes or acts for another.
indicatorformal /ˈɪndəˌkeɪtər/ something measurable that points to a larger condition Synonyms: sign, measure, marker	a sign that can be measured	Infant mortality remains the most widely used indicator of a country's health system.
It points to something rather than capturing it whole, so writers speak of a good or a crude indicator, and the preposition is 'of'. In statistics an 'indicator variable' simply marks yes or no.
indexformal /ˈɪndɛks/ a single number built from several measures to allow comparison Synonyms: combined score, composite measure, scale	a combined score	The index combines income, schooling and life expectancy into one score per country.
Two plurals split by sense: indexes for lists at the back of books, indices, /ˈɪndəsiz/, for the mathematical and economic sense. You construct, compute or revise an index; the parts are weighted before being added.
time seriesformal /ˈtaɪm ˌsɪriz/ a set of measurements of the same thing taken at regular times Synonyms: a run of dated figures, dated sequence	values over successive periods	The time series runs from 1960 to 2015, with one observation for each quarter.
Series is already singular here, so write 'a time series is' and 'two time series are'; there is no form 'serie'. The order of the points carries the information, which is why shuffling them destroys the data. Also used before a noun: 'time series analysis'.
incidenceformal /ˈɪnsədəns/ how often new cases of something appear in a period Synonyms: occurrence, frequency, new cases per period	rate of new cases	The incidence of new infections fell sharply in the three months after the campaign.
Incidence counts new cases over a period while prevalence counts all existing cases at one moment, a distinction that decides which number answers your question. The frame is 'the incidence of X'. Do not confuse it with 'incident'.
prevalenceformal /ˈprɛvələns/ how widespread something is at a given time Synonyms: how common something is, share, extent	share of existing cases	Prevalence of the condition among adults is estimated at four to six percent.
It captures a snapshot of all existing cases, so a long-lasting condition can have high prevalence but low incidence. The frame is 'the prevalence of X among Y'. The adjective 'prevalent' means widespread, not powerful.
empiricalformal /ɛmˈpɪrɪkəl/ based on what has been observed or measured rather than on ideas alone Synonyms: observed, fact-based, experience-based	based on observed facts	The argument is elegant, but empirical support for it remains thin.
Its opposite is theoretical, not false: an empirical claim can be wrong and a theoretical one right. Frames: 'empirical support', 'empirical work', 'an empirical question'. The adverb 'empirically' means through observation.
anecdotalformal /ˌænɪkˈdoʊtəl/ based on a few personal stories rather than systematic study Synonyms: story-based, informal, unsystematic	based on single stories	Support for the method is largely anecdotal, resting on reports from a handful of teachers.
Nearly always faintly critical, and the fixed phrase is 'anecdotal evidence'. It does not mean untrue; it means too unsystematic to bear weight. The polite contrast is 'anecdotal but suggestive', which concedes and hedges at once.
to triangulateformal /traɪˈæŋɡjəleɪt/ check a finding by approaching it from more than one direction Synonyms: cross-check, confirm from several sources	check by several routes	The team triangulated the interview accounts against attendance records and school reports.
The frame is 'triangulate X against Y' or 'triangulate across sources'. The noun triangulation is common in mixed methods writing. The point is that three weak views may be stronger together than any one alone.
to compileformal /kəmˈpaɪl/ gather information from several places into one collection Synonyms: gather, put together, assemble	gather into one collection	The figures were compiled from national registers covering the years 1990 to 2015.
'Compile X from Y' names the sources, and the passive is standard: 'were compiled from'. It suggests patient gathering and little change to the content. A separate computing sense is turning written code into a runnable program.
causalityformal /kɔˈzæləti/ the relation between an event and the thing that makes it happen Synonyms: cause and effect, causation	one thing causing another	Establishing causality needs more than two lines on a graph moving together.
Causation is the commoner word for the same idea, while causality is preferred where the logic itself is under discussion. The adjective is causal, /ˈkɔzəl/, which both learners and spellcheckers confuse with casual. Abstract and uncountable, so there is no plural.
estimateneutral /ˈɛstəmət/ a figure worked out from partial information rather than counted exactly Synonyms: calculated figure, approximation, projection	a calculated approximate figure	Published estimates of the cost range from two to nine billion dollars.
The noun ends /-ət/ and the verb ends /-eɪt/, a stress-free but audible difference. Countable: 'a conservative estimate', 'estimates range from'. Conservative here means deliberately cautious, chosen so the claim is hard to attack.
to normalizeformal /ˈnɔrməˌlaɪz/ change numbers to a common scale so they can be compared fairly Synonyms: rescale, adjust, standardize	put on a common scale	Counts were normalized by population so that large and small cities could be compared.
The frame is 'normalize X by Y', naming what the numbers are divided by. A second and quite different sense is to make something socially accepted, as in 'normalizing overwork'. British spelling: normalise.
field notesneutral /ˈfild ˌnoʊts/ written records made at the time of watching people or places Synonyms: observation records, on-site notes	notes written while observing	Her field notes describe the waiting room in more detail than any of the interviews do.
Writing them up the same day is the rule, since the value lies in what memory would lose by evening. They count as data in their own right and can be quoted like an interview. Always plural, and two words in American usage.
causalformal /ˈkɔzəl/ having to do with one thing making another happen Synonyms: cause-related, explanatory	to do with cause	The study shows a clear pattern but stops short of any causal claim.
Used before a noun: 'a causal claim', 'causal inference', 'a causal link'. Keep it apart from 'casual', which differs by one letter and means relaxed. The adverb 'causally' appears in 'causally related'.
causationformal /kɔˈzeɪʃən/ the fact of one thing actually making another happen Synonyms: cause, production of an effect	one thing producing another	Showing that two measures move together is far easier than showing causation.
Normally set against correlation in the standard warning that one does not establish the other, and writers 'establish', 'infer' or 'rule out' causation. 'Reverse causation' names the possibility that the effect came first.
to attributeformal /əˈtrɪbjut/ say that something was caused by or produced by a particular thing Synonyms: credit to, put down to, ascribe	say what caused it	The authors attribute the improvement to smaller classes rather than to the new textbook.
The frame is 'attribute X to Y', effect first and cause second, and it cannot be reversed. The verb stresses the second syllable while the noun 'attribute', a quality, stresses the first. Very common in the passive.
attributableformal /əˈtrɪbjətəbəl/ able to be explained as caused by something in particular Synonyms: traceable to, caused by, due to	able to be blamed on	About a fifth of the fall in deaths is attributable to earlier screening.
Always used with 'to' after it, and typically with a share or proportion in front: 'the share attributable to smoking'. It hedges more carefully than 'caused by', since it claims only that the accounting works out.
to result inneutral /rɪˈzʌlt ɪn/ have something as its effect Synonyms: lead to, produce, bring on	have as its effect	Higher tuition did not result in fewer applications from the lowest income group.
'Result in' points forward to the effect while 'result from' points back to the cause, so the two prepositions run in opposite directions. Both take a noun or an -ing form, never a that-clause.
to result fromneutral /rɪˈzʌlt frʌm/ happen because of something Synonyms: arise from, come from, follow from	be caused by	The gap appears to result from differences in teaching time rather than class size.
The cause follows 'from'. Writers soften it with 'appears to result from' or 'may result from', which is the most common hedge in a discussion section. Do not write 'resulted of' or 'resulted by'.
to precipitateformal /prɪˈsɪpəteɪt/ make something happen suddenly or earlier than expected Synonyms: set off, bring on, hasten	trigger suddenly	The loss of two major funders precipitated the closure of the program in 2009.
It adds suddenness to plain causing, and its objects are usually unwelcome: a crisis, a collapse, a resignation. The verb ends /-eɪt/ and the adjective 'precipitate', meaning hasty, ends /-ət/. A chemical sense also exists.
antecedentformal /ˌæntəˈsidənt/ something that comes before and may help explain what follows Synonyms: forerunner, earlier cause, precursor	something that comes before	The paper traces the economic antecedents of the reforms back to the 1970s.
It claims order in time and only hints at cause, which makes it the careful choice when you cannot show more. Usually plural and followed by 'of'. In grammar it also names the noun a later pronoun refers back to.
determinantformal /dɪˈtɜrmənənt/ something that strongly decides how another thing turns out Synonyms: deciding factor, driver, influence	a deciding factor	Parental education remains the strongest determinant of school success in these data.
It is a strong word, claiming that the thing largely settles the outcome, so writers qualify it with 'a major' or 'the strongest'. Followed by 'of': 'the determinants of health'. The adjective 'determinate' means fixed.
predictorformal /prɪˈdɪktər/ something that helps forecast a later result without necessarily causing it Synonyms: forecasting measure, indicator, sign	a measure that forecasts	Early vocabulary size is a strong predictor of reading ability at age nine.
Deliberately weaker than cause, which is why it is safe in a results section: 'a strong predictor' comments on how well something forecasts, not on how it works. The preposition is 'of': 'a predictor of dropout'.
confounderformal /kənˈfaʊndər/ something that affects both of two things and makes them look connected Synonyms: a third factor, lurking variable	a hidden third cause	Age is an obvious confounder: older patients both smoke less and die sooner.
Also written confounding variable, and the verb is confound: 'age confounds the comparison'. It has to be linked to both sides, so something affecting only the outcome is not one. Controlling for it, or matching on it, is the standard answer.
to predisposeformal /ˌpridɪˈspoʊz/ make someone or something more likely to be affected in a certain way Synonyms: make liable, incline, expose	make more likely to	Early stress may predispose children to difficulties with attention later on.
'Predispose X to Y', and the outcome is nearly always unwanted. It claims raised risk rather than certainty, which is why it sits comfortably with 'may'. The noun is predisposition, as in 'a genetic predisposition to'.
to underlieformal /ˌʌndərˈlaɪ/ be the hidden cause or basis of something Synonyms: be the basis of, cause, support	lie beneath as a cause	The same mechanism may underlie both effects, though neither study tested it directly.
Takes a direct object with no preposition, and the participle frame 'the processes underlying X' is the one to learn. Forms: underlies, underlay, underlain. It claims something deeper rather than something earlier, which is why it suits mechanisms.
to presupposeformal /ˌprisəˈpoʊz/ treat something as true before the argument begins Synonyms: assume, take for granted, take as given	take for granted, unargued	This reading presupposes that the letters were written in the order they were filed.
Followed by a that-clause or a noun. It is often mildly critical, pointing out that the assumption was never argued for. The noun is presupposition. Compare 'assume', which is neutral, and 'presume', which claims more confidence.
deductiveformal /dɪˈdʌktɪv/ reasoning from a general rule down to a particular case Synonyms: rule-based, top-down, logical	reasoning from general rules	The chapter takes a deductive approach, starting from three principles and testing each one.
Used before a noun: 'a deductive argument', 'deductive reasoning'. A valid deductive argument guarantees its conclusion if the premises hold, which inductive reasoning never does. Its partner is inductive, and the noun is deduction.
inductiveformal /ɪnˈdʌktɪv/ reasoning from particular cases up to a general pattern Synonyms: bottom-up, case-based, exploratory	reasoning from cases up	Codes were developed inductively from the interviews rather than fixed in advance.
Frames: 'an inductive approach', 'coded inductively'. Its conclusions are always open to the next case, so inductive claims are stated with hedges. The noun is induction, which in a different field means causing something to start.
to hypothesizeformal /haɪˈpɑθəˌsaɪz/ suggest an idea as something to be tested Synonyms: propose, suggest, theorize	propose as a testable idea	We hypothesized that feedback delivered within a day would produce the largest gain.
Takes a that-clause, and in American papers it usually appears in the past tense to show the idea came before the results. British spelling is 'hypothesise'. It claims only that the idea is worth testing.
null hypothesisformal /ˌnʌl haɪˈpɑθəsɪs/ the starting assumption that there is no real difference or effect Synonyms: no-effect assumption, baseline assumption	the no-effect assumption	We can reject the null hypothesis of no difference between the two teaching methods.
The fixed verbs are 'reject' and 'fail to reject', and the second is deliberate: nobody accepts a null hypothesis, they only fail to knock it down. The plural is 'null hypotheses', /haɪˈpɑθəsiz/.
sampling frameformal /ˈsæmplɪŋ ˌfreɪm/ the list of all the cases a sample could be drawn from Synonyms: the source list, population list	the list you sample from	The sampling frame was the register of licensed drivers, which leaves out anyone who never applied.
Whoever is absent from it can never appear in the sample, so the frame sets the real limit on who a study is about. Naming it is expected in a method section, and criticism usually starts there. Countable: 'two overlapping sampling frames'.
to postulateformal /ˈpɑstʃəleɪt/ assume something in order to build an explanation on it Synonyms: assume, propose, posit	assume for the argument	The model postulates a fixed limit on how much can be held in mind at once.
Models and theories are the usual subjects: 'the theory postulates'. It signals that the thing assumed cannot be observed directly. The noun is 'a postulate', /ˈpɑstʃələt/, with a weaker final vowel than the verb.
to argueneutral /ˈɑrɡju/ give reasons for a position you want the reader to accept Synonyms: make the case, contend, hold	make a case in writing	In this article I argue that the decline began well before the reforms of 1994.
In academic writing it means to reason for a position, not to quarrel, and it takes a that-clause: 'we argue that'. 'Argue for' plus a noun also works. It is the standard verb for announcing your own claim.
tenableformal /ˈtɛnəbəl/ able to be defended against criticism Synonyms: defensible, arguable, sound	able to be defended	After three failed replications, the original explanation is no longer tenable.
Most often in the negative: 'no longer tenable', 'hardly tenable', or the opposite form 'untenable'. It judges whether a position can be held, not whether it happens to be true. Positions and views are tenable; people are not.
assertionformal /əˈsɜrʃən/ a statement made firmly, often without support given Synonyms: claim, statement, declaration	a firm unproved statement	The assertion that class size does not matter is hard to square with these numbers.
Calling something 'a mere assertion' is a standard way of objecting that no support was offered. Writers 'make' an assertion and 'challenge' or 'test' one, and it is often followed by 'that'.
sensitivity analysisformal /ˌsɛnsəˈtɪvəti əˈnæləsɪs/ a repeat of an analysis with different choices, to see if the answer changes Synonyms: a robustness check, what-if test	rerunning with other choices	A sensitivity analysis showed the finding held whether or not the two outliers were dropped.
The point is fragility: if a conclusion survives only one particular set of decisions, this is what exposes it. The frames are 'in a sensitivity analysis' and 'the results were not sensitive to'. Plural analyses, ending /siz/.
to suggestneutral /səɡˈdʒɛst/ point toward a conclusion without proving it Synonyms: indicate, point to, imply	point toward, weakly	These results suggest that practice spread over a week is more effective than one long session.
The single most useful claim verb in a results section, because it is deliberately weak: 'the data suggest that' concedes room for doubt. Use 'suggest that' plus a clause; 'suggest us' and 'suggest me' are the classic errors.
to indicateneutral /ˈɪndəkeɪt/ show or point to something without stating it fully Synonyms: show, point to, reveal	point to, moderately	Table 3 indicates that response times rose steadily across the four blocks.
Sits between 'suggest' and 'show' in strength, and takes a that-clause or a noun. Tables, figures and results are its usual subjects. It reports what can be read off the material rather than what follows from it.
subgroup analysisformal /ˈsʌbˌɡrup əˈnæləsɪs/ a check of whether a result holds inside smaller parts of the sample Synonyms: a split-sample check, group-by-group test	testing within smaller groups	The subgroup analysis found the benefit only among women over fifty, which needs replication.
Many small comparisons mean many chances to find something by luck, so a subgroup result counts as a lead rather than a finding unless it was planned in advance. The plural is subgroup analyses, /əˈnæləsiz/. Writers signal honesty with 'prespecified' or 'exploratory'.
thematic analysisformal /θəˈmætɪk əˈnæləsɪs/ a way of studying written or spoken material by grouping it into themes Synonyms: theme-based analysis, pattern coding	finding patterns in text	The thematic analysis produced four themes, each illustrated with two or three quotations.
A theme is more than a topic that came up: it has to say something, which is why good papers word their themes as short claims. The plural is analyses, /əˈnæləsiz/. Reviewers ask who coded, how many coded, and whether the themes were fixed in advance.
to validateformal /ˈvæləˌdeɪt/ check that something works or measures what it should Synonyms: check, confirm, test	check that it works	The questionnaire was validated on a separate group of two hundred readers.
In research it means testing a measure against something independent, and the frame is 'validate X on' or 'against Y'. The everyday sense of making someone feel accepted belongs to a different register. Noun: validation.
to falsifyformal /ˈfɔlsəˌfaɪ/ show by testing that an idea is false Synonyms: disprove, refute, test to destruction	show to be false	A theory that no observation could falsify is of little scientific use.
Two senses pull in opposite directions: in the philosophy of science it means to show an idea false, while elsewhere it means to alter records dishonestly. Context decides. The adjective 'falsifiable' belongs only to the first sense.
spuriousformal /ˈspjʊriəs/ looking like a real link but caused by something else Synonyms: false, misleading, apparent	seeming real but false	The link between ice cream sales and drowning is spurious; both rise with temperature.
The standard phrase is 'a spurious correlation', where a third thing drives both measures. It marks a relationship as an artifact rather than a discovery. A broader sense, 'spurious reasoning', means false or fake.
reciprocalformal /rɪˈsɪprəkəl/ working in both directions between two things Synonyms: two-way, mutual, both-way	working both ways	The relationship appears reciprocal: reading builds vocabulary, and vocabulary encourages reading.
Frames: 'a reciprocal relationship', 'reciprocal influence'. It is the useful answer when neither direction of cause can be ruled out. The adverb 'reciprocally' is rare; writers prefer 'in both directions'.
inverseformal /ˈɪnvɜrs/ moving in the opposite direction, so that one rises as the other falls Synonyms: opposite, reversed, negative	opposite in direction	There is an inverse relationship between household income and hours spent commuting.
The fixed phrase is 'an inverse relationship between X and Y', and 'inversely related' works the same way. It describes direction only and says nothing about strength. As an adjective it stresses the first syllable.
weighted averageformal /ˈweɪtəd ˈævərɪdʒ/ an average in which some values count for more than others Synonyms: an average by size, adjusted mean	an average with unequal shares	The national figure is a weighted average, so the larger states move it much more.
The weights say how much each value counts, usually the size of the group behind it. Giving one without saying what the weights were is a standard criticism. In running prose 'weighted by population' does the same work as the noun phrase.
counterfactualformal /ˌkaʊntərˈfæktʃuəl/ a statement about what would have happened if things had been different Synonyms: the what-if case, comparison case	what would have happened	Any claim about the policy's effect needs a counterfactual: what these towns would have done anyway.
Causal claims rest on it, since an effect is the gap between what happened and what would have happened without the cause. Works as both a noun and an adjective: 'the counterfactual', 'counterfactual reasoning'. A control group is one way of standing in for it.
to hinge onneutral /ˈhɪndʒ ɑn/ depend completely on one thing Synonyms: depend on, turn on, rest on	depend entirely on	The whole argument hinges on whether the two samples are really comparable.
Stronger than 'depend on' because it names a single decisive point, and it is often followed by 'whether'. Its subjects are arguments, conclusions and decisions. 'Hinge upon' is the more formal variant.
self-reportneutral /ˌsɛlf rɪˈpɔrt/ information given by people about their own behavior or feelings Synonyms: own account, self-description	what people say about themselves	Everything here is self-report, so the figures show what people say they do.
It is the cheapest way to collect almost anything and the easiest to challenge, so careful papers add a second source where they can. Used as a noun and before another noun: 'self-report data', 'a self-report measure'. The verb keeps the hyphen: participants self-reported their weight.
to equateformal /ɪˈkweɪt/ treat one thing as equal to or the same as another Synonyms: identify with, treat as equal, liken	treat as the same	It is a mistake to equate a high test score with deep understanding.
'Equate X with Y', not 'to Y', in careful American usage. It often appears in a warning: 'we should not equate'. Compare 'conflate', which says the two were merged by mistake rather than on purpose.
to disaggregateformal /dɪsˈæɡrəˌɡeɪt/ separate a combined figure into the smaller groups that make it up Synonyms: break down, separate out	break a total into parts	Once we disaggregate the results by age, the gap between the two regions disappears.
The frame is 'disaggregate X by Y', where Y is the grouping: by sex, by region, by year. Its opposite is aggregate, and the pair belongs together because a claim can hold in the total and fail in every group inside it. The noun is disaggregation.
to differ fromneutral /ˈdɪfər frʌm/ be unlike something else in some way Synonyms: vary from, contrast with, be distinct from	be unlike	Adult learners differ from children in the strategies they bring to a new language.
The preposition is 'from', and 'differ than' is an error. 'Differ in' names the respect: 'the groups differed in age but not in schooling'. 'Differ with someone' means to disagree, which is a separate use.
to varyneutral /ˈvɛri/ be different in different cases, places or times Synonyms: differ, range, fluctuate	change across cases	Class sizes varied widely across the twelve schools in the sample.
Common frames are 'vary across', 'vary by' and 'vary with': across regions, by age group, with temperature. 'Varies from X to Y' gives a range. The noun 'variation' is countable; 'variability' names the tendency itself.
to pertain toformal /pərˈteɪn tu/ be about or apply to something in particular Synonyms: concern, bear on, have to do with	relate to, apply to	Only the first three questions pertain to the period after the reform.
Always with 'to', and distinctly formal; in speech people say 'be about'. It is useful for marking the limits of what a source covers. The adjective 'pertinent' is the everyday partner and means directly relevant.
with respect toformal /wɪð rɪˈspɛkt tu/ used to name the exact point being compared or discussed Synonyms: regarding, as for, concerning	as regards, concerning	The two groups were matched with respect to age, schooling and reading level.
A fixed phrase followed by a noun, common where the writer must be exact about the dimension compared: matched, similar or different with respect to something. 'With regard to' is interchangeable; 'in respect of' is British and legal.
to qualify (a claim)formal /ˈkwɑləˌfaɪ/ add something to a statement so that it claims less Synonyms: limit, soften, restrict	narrow what you claimed	The authors qualify their claim by noting that all three schools were unusually well funded.
Its objects are claims, statements and conclusions: 'the authors qualify this conclusion'. Qualifying narrows a claim rather than improving it, and doing it in the same sentence as the claim is normal practice. The noun is qualification, often plural.
skewedneutral /skjud/ having more values at one end than the other, so the average misleads Synonyms: lopsided, uneven, distorted	pulled to one side	Income data are almost always skewed, which is why the median is reported instead of the mean.
Two uses sit close together: a skewed distribution leans to one side, while skewed results have been bent by something. Data are skewed to the right when a long tail of high values pulls the mean upward. The noun is skew, and the frame is 'skewed toward'.
purportedformal /pərˈpɔrtɪd/ said by others to be true, without the writer agreeing Synonyms: alleged, claimed, so-called	claimed but unproven	The purported benefits of the supplement rest on two small unpublished studies.
Used before a noun and always distancing: it reports a claim while declining to share it. The related verb frame is 'X purports to show Y', which carries the same doubt. Do not use it for claims you accept.
putativeformal /ˈpjutətɪv/ generally believed to be the thing named, without being proven Synonyms: supposed, assumed, presumed	assumed to be so	The putative cause is a virus, but no sample has yet been recovered.
Used only before a noun: 'the putative mechanism', 'a putative ancestor'. It is milder than 'purported', reporting a working assumption rather than doubting a claim. Common in scientific writing about things not yet observed.
response rateneutral /rɪˈspɑns ˌreɪt/ the proportion of the people asked who actually took part Synonyms: share who replied, participation share	the share who answered	A response rate of thirty percent leaves a large question about the seventy who stayed silent.
Given as a percentage and expected in any survey paper, because who is missing matters more than how many. A low rate is not fatal if those who stayed away resemble the rest, which is hard to show. Achieve and report are the verbs it takes.
comparativelyformal /kəmˈpærətɪvli/ when measured against something similar Synonyms: relatively, rather, fairly	by comparison	Comparatively little work has looked at how these habits form in adulthood.
Near-identical to 'relatively' in use, but it points more openly to a comparison with a named field or period. The set frame 'comparatively little work' is a standard way to open a gap in the literature.
moderateformal /ˈmɑdərət/ neither large nor small in size or strength Synonyms: medium, middling, fair	medium in degree	Agreement between the two raters was moderate rather than strong.
As an adjective it ends /-ət/; the verb 'to moderate', meaning to make weaker or to lead a discussion, ends /-eɪt/. Statistical conventions often define weak, moderate and strong ranges, so the word can be surprisingly precise.
dispersionformal /dɪˈspɜrʒən/ the amount by which values in a set are spread away from the middle Synonyms: spread, scatter	how spread out values are	Two groups can share a mean and still differ sharply in dispersion.
Where the mean says where the middle is, dispersion says how far the cases sit from it, and standard deviation is the usual way of putting a number on that. Uncountable in this sense. The everyday physical sense of scattering is why 'spread' reads more plainly in speech.
appreciableformal /əˈpriʃəbəl/ large enough to be noticed or measured Synonyms: noticeable, measurable, considerable	big enough to notice	No appreciable difference emerged between the morning and afternoon sessions.
Most often negative: 'no appreciable difference', 'without appreciable change'. It concerns being noticeable rather than being important, and it has nothing to do with feeling grateful, which is 'appreciative'.
discernibleformal /dɪˈsɜrnəbəl/ able to be seen or noticed, if only just Synonyms: noticeable, detectable, visible	just able to be seen	There was no discernible pattern in which families chose to leave the study.
Overwhelmingly used in the negative: 'no discernible effect', 'no discernible trend'. It reports the limits of what the observer could see, which is a more modest claim than saying nothing was there. Verb: to discern.
preliminaryformal /prɪˈlɪməˌnɛri/ done or reported early, before the full work is finished Synonyms: early, initial, first-stage	early and incomplete	These preliminary results come from the first sixty participants only.
Used before a noun: 'preliminary findings', 'a preliminary analysis'. It buys the writer room by promising that fuller work follows, so do not use it for work that is actually complete. The plural noun 'preliminaries' means opening steps.
equivocalformal /ɪˈkwɪvəkəl/ open to more than one reading, so it settles nothing Synonyms: ambiguous, mixed, unclear	open to two readings	Results across the six sites were equivocal, with three positive and three negative.
Slightly stronger than 'inconclusive': it says the material actively points in two directions rather than simply falling short. Its opposite 'unequivocal' means admitting no doubt. It describes findings and language, not people.
suggestiveformal /səɡˈdʒɛstɪv/ pointing toward a conclusion without establishing it Synonyms: indicative, hinting, pointing to	hinting at a conclusion	The timing is suggestive of a shared cause, though other readings remain open.
The academic frame is 'suggestive of' plus a noun, and 'suggestive but not conclusive' is a standing pair. Note that in general English 'suggestive' can hint at something indecent, so the 'of' phrase keeps the academic sense clear.
unsubstantiatedformal /ˌʌnsəbˈstænʃieɪtɪd/ not supported by anything that can be checked Synonyms: unsupported, unproven, undocumented	offered without support	The claim of a fivefold benefit is unsubstantiated in the published record.
Applies to claims, reports and allegations, not to people. It says support is missing, not that the claim is false, which makes it a careful criticism. The verb behind it is 'substantiate'.
dubiousformal /ˈdubiəs/ likely to be wrong, or of doubtful quality Synonyms: doubtful, questionable, shaky	shaky; hard to trust	The comparison rests on dubious assumptions about how the two tests are scored.
Two directions: 'dubious assumptions' means the things themselves are doubtful, while 'dubious about the method' means the person has doubts. The first is the academic use. It is franker than 'questionable' and edges toward disapproval.
questionableformal /ˈkwɛstʃənəbəl/ not certainly correct or acceptable Synonyms: doubtful, debatable, uncertain	open to doubt	Whether these results apply to older workers is questionable at best.
It reports doubt about correctness and, with words like 'practices' or 'conduct', edges toward doubt about honesty. Politer than 'wrong' and firmer than 'debatable'. Common frames: 'a questionable assumption', 'questionable at best'.
to call into questionformal /ˈkɔl ˌɪntu ˈkwɛstʃən/ give reasons to doubt something previously accepted Synonyms: challenge, undercut, dispute	cast doubt on	These findings call into question the standard explanation of the effect.
A fixed phrase; the object normally follows the whole phrase, though a short pronoun can sit inside it. Its subjects are findings, results and arguments rather than people, which keeps the criticism impersonal.
to challengeneutral /ˈtʃælɪndʒ/ question whether something accepted is really right Synonyms: dispute, question, contest	dispute openly	Two recent papers challenge the assumption that the two skills develop together.
Takes a direct object with no preposition: 'challenge the view', not 'challenge to the view'. In academic prose it means to dispute rather than to invite a contest. As a noun 'a challenge to X' takes 'to'.
measurement errorformal /ˈmɛʒərmənt ˈɛrər/ the difference between a recorded value and the true one Synonyms: inaccuracy, recording error	the gap from the true value	Self-reported height carries measurement error in one direction: people round upward.
Not a mistake in the ordinary sense, since every instrument has some; the question is whether it is random or leans one way. Random error widens the spread, while systematic error shifts everything together. Uncountable when it names the problem in general.
ceiling effectformal /ˈsilɪŋ ɪˌfɛkt/ a limit in a test that hides differences between the strongest cases Synonyms: upper limit on scores, a test too easy	no room to score higher	A ceiling effect is likely here, since nearly every student answered all twenty items correctly.
Named for the ceiling it hits: scores pile up at the maximum and real differences above that point cannot show. Its mirror image is a floor effect, where a task is too hard for anyone to differ. Reported as 'a ceiling effect on the vocabulary measure'.
to confineformal /kənˈfaɪn/ keep something within stated limits Synonyms: restrict, limit, keep to	keep within limits	We confine our attention to studies published in English after 1990.
The frame is 'confine X to Y', and the passive 'is confined to' is very common in statements of scope. It announces a limit the writer has chosen, in contrast to a constraint imposed from outside. Noun: confines, meaning limits.
external validityformal /ɪkˈstɜrnəl vəˈlɪdəti/ how well a result holds outside the people and setting a study used Synonyms: generalizability, reach beyond the sample	how far results carry	External validity is the weak point here, since every participant was a first-year student.
Set against internal validity, which asks whether the result is even true of the study itself: a tightly controlled experiment can be strong on one and weak on the other. Threaten and limit are the verbs it takes. Reviewers raise it whenever a sample looks narrow.
to generalizeformal /ˈdʒɛnərəˌlaɪz/ apply a finding beyond the cases actually studied Synonyms: extend, apply widely, broaden	extend beyond the cases studied	It is unclear how far these results generalize to schools outside large cities.
The academic frame is 'generalize to' a wider group, and 'generalize from' the cases in hand. The noun 'generalizability' names how far a finding travels. In everyday use 'generalize' means to speak too loosely, a faintly negative sense.
cautiousneutral /ˈkɔʃəs/ careful to avoid claiming more than is safe Synonyms: careful, guarded, restrained	careful, not overclaiming	A cautious reading of these numbers would put the gain at two points, not five.
In academic writing it is praise, not criticism: 'a cautious interpretation' is a safe one. Frames: 'cautious about', 'cautiously optimistic'. The noun is caution, and 'with caution' is the standard adverbial phrase.
to cautionformal /ˈkɔʃən/ warn someone that something may be wrong or risky Synonyms: warn, advise against, urge care	warn against	The authors caution against reading the drop as a lasting improvement.
Two frames dominate: 'caution against' plus an -ing form, and 'caution that' plus a clause. It is the standard verb for a discussion section limiting how a result should be read. The related noun 'a note of caution' opens such a passage.
flawedformal /flɔd/ containing a fault serious enough to weaken the result Synonyms: faulty, defective, unsound	spoiled by a fault	Two of the four trials were seriously flawed and are excluded here.
A strong criticism: it says something inside the work spoils it, not that the topic was hard. Usual pairings are 'deeply', 'seriously' and 'fatally flawed'. The noun 'a flaw in the design' names the specific fault.
rigorousformal /ˈrɪɡərəs/ carried out with great care and strict standards Synonyms: strict, thorough, exacting	strictly careful	This is the most rigorous test of the idea attempted so far.
High praise in academic writing: 'a rigorous test', 'rigorous standards', 'methodologically rigorous'. The noun is rigor, spelled rigour in British English. It describes the care taken, not how difficult the work was.
preregistrationformal /ˌpriˌrɛdʒəˈstreɪʃən/ the act of recording a study's plan publicly before the data are collected Synonyms: an advance plan on record, prior registration	filing the plan in advance	Preregistration makes it harder to present a chance finding as something the study set out to test.
The verb is preregister and such a study is preregistered; the document itself can also be called a preregistration. Written solid in American usage, with no hyphen. It separates what was predicted from what was noticed afterward, which is the whole point of it.
to referenceformal /ˈrɛfərəns/ name the source of an idea in the correct form Synonyms: cite, credit, source	name your source	Older papers are referenced in full the first time they appear.
Used as a verb mainly in writing about writing, and usually passive: 'sources are referenced'. As a noun 'a reference' is the entry in the list at the end. 'With reference to' is a fixed phrase meaning concerning.
attributionformal /ˌætrəˈbjuʃən/ the act of saying who or what something came from Synonyms: credit, acknowledgment, ascription	naming the source or cause	The same passage appears without attribution in at least two later textbooks.
A second sense covers naming a cause, as in 'the attribution of the decline to policy', so the two uses sit far apart. In the citing sense it has no plural: 'without attribution', 'proper attribution'. The verb for both is 'attribute'.
plagiarismformal /ˈpleɪdʒəˌrɪzəm/ using someone else's words or ideas without saying where they came from Synonyms: copying, theft of ideas, unacknowledged borrowing	using others' work as yours	The journal withdrew the paper after two clear cases of plagiarism were found.
Usually uncountable, though 'cases of plagiarism' is countable. It covers ideas as well as wording, and it does not require intention, which is why careless note-taking is a real risk. The verb is 'plagiarize'.
retractionformal /rɪˈtrækʃən/ the formal withdrawal of a published paper by its authors or the journal Synonyms: withdrawal, formal removal, cancellation	formal withdrawal of a paper	The retraction notice appeared two years after the paper was first published.
A retraction pulls a paper from the record while leaving it visible and marked, so later work can still cite it as retracted. The fixed phrase is 'a retraction notice', and the verb is 'retract'.
extantformal /ˈɛkstənt/ still existing or available today Synonyms: surviving, existing, remaining	still in existence	Only three copies of the 1690 edition are extant, all in private collections.
Works before a noun and after 'be': 'extant manuscripts', 'no copies are extant'. The standing phrase 'the extant literature' means the body of published work as it now stands. Its opposite for species is 'extinct'.
replicationformal /ˌrɛpləˈkeɪʃən/ the act of running a study again to see if the result appears once more Synonyms: a repeat study, independent repetition	repeating a study	Two attempts at replication produced nothing like the effect reported in the first paper.
A direct replication copies the method as closely as it can, while a conceptual one tests the same idea another way. Countable when it names a particular attempt: 'three replications failed'. The verb is replicate, and a finding that survives is said to replicate.
to dateformal /tə ˈdeɪt/ up to the present time Synonyms: so far, as yet, thus far	up to now	No study to date has followed these children beyond the age of twelve.
A fixed phrase that follows the noun it limits: 'the largest trial to date', 'work to date'. It pairs with the present perfect, since it measures from the past up to now. Do not confuse it with the verb 'to date something'.
literatureformal /ˈlɪtərətʃər/ all the published work written on a particular subject Synonyms: published work, research record, scholarship	published work on a topic	The literature on bilingual reading has grown quickly over the past decade.
In this use it has nothing to do with novels and poetry: 'the literature on X' means everything published on that topic. Almost always with 'the' plus 'on', and there is no plural. Frames: 'a gap in the literature', 'this literature suggests'.
to reconcileformal /ˈrɛkənˌsaɪl/ explain how two facts that seem to disagree can both be true Synonyms: square, make consistent, bring together	make two findings fit	It is hard to reconcile these results with the null finding reported in 2014.
'Reconcile X with Y' is the frame, and the difficulty is usually admitted along the way: 'hard to reconcile', 'cannot easily be reconciled'. In this sense the things reconciled are findings or accounts rather than people. Noun: reconciliation.
scholarlyformal /ˈskɑlərli/ belonging to serious academic study, and written for that audience Synonyms: academic, learned, research-based	academic in standard	The claim has never appeared in a scholarly journal, only in the press.
An adjective despite the -ly ending, so there is no adverb form; write 'in a scholarly manner'. Frames: 'scholarly work', 'scholarly consensus', 'a scholarly audience'. Of a person it means learned and careful.
ethnographyformal /ɛθˈnɑɡrəfi/ a detailed written study of a group based on long close observation Synonyms: field study, cultural description	study by living among	Her ethnography of a hospital night shift took fourteen months of observation to complete.
It names both the method and the book that comes out of it, so 'an ethnography' is countable while ethnography as an approach is not. The adjective is ethnographic and the researcher an ethnographer. Time in the field is what the word promises; a week of interviews is not one.
interdisciplinaryformal /ˌɪntərˈdɪsəpləˌnɛri/ combining the methods or ideas of more than one field Synonyms: cross-field, multi-field, joint	across several fields	The project is interdisciplinary, bringing together historians, statisticians and public health staff.
Written as one word with no hyphen after 'inter'. It claims that the fields genuinely work together, where 'multidisciplinary' only requires that several are present. Frames: 'interdisciplinary research', 'an interdisciplinary team'.
manuscriptformal /ˈmænjəˌskrɪpt/ a piece of writing sent for publication but not yet printed Synonyms: paper, draft, submission	an unpublished paper	The manuscript was returned twice before the editors accepted it.
In publishing it means a text awaiting a decision, while in history it means a document written by hand. You submit, revise and resubmit a manuscript, abbreviated MS or ms. in notes.
to submitformal /səbˈmɪt/ send work to be considered for publication or approval Synonyms: send in, hand in, put forward	send in for judgment	The paper was submitted in March and accepted eight months later.
Double the t before endings: submitted, submitting. 'Submit X to Y' names the journal or committee. A second, formal sense introduces an argument: 'we submit that the evidence points elsewhere'.
reviewerformal /rɪˈvjuər/ an expert who reads a paper before publication and advises on it Synonyms: referee, assessor, expert reader	an expert who judges a paper	One reviewer asked for an extra analysis; the other recommended rejection.
A reviewer advises and the editor decides, and confusing the two misreads a rejection letter. Reviewers are usually anonymous, which is why authors write 'we thank the reviewers'. 'Referee' means the same in the sciences.
avenue (for research)formal /ˈævəˌnu/ a possible direction that later work could take Synonyms: line of inquiry, direction, path	a possible line of inquiry	One promising avenue is whether the effect survives in much older learners.
The standing phrase is 'avenues for future research', which closes many discussion sections, and you open, suggest or pursue an avenue. The plural is the norm, since offering only one can read as an admission of narrowness.
authorshipformal /ˈɔθərˌʃɪp/ the fact of being named as a writer of a published work Synonyms: credit, byline, writing credit	credit for writing	Authorship was decided by how much each person contributed to the design.
It covers both who gets named and in what order, which in many fields carries real weight. Frames: 'questions of authorship', 'joint authorship'. 'Disputed authorship' is also used of old texts whose writer is unknown.
affiliationformal /əˌfɪliˈeɪʃən/ the university or organization a person officially belongs to Synonyms: institution, connection, attachment	the institution you belong to	Her affiliation is given as the second institution on the title page.
On a paper it is the line under each author's name, and 'affiliation with' names the body: 'an affiliation with the Institute'. The verb is 'be affiliated with', and 'affiliations' can also mean political loyalties.
in pressformal /ɪn ˈprɛs/ accepted for publication but not yet printed Synonyms: forthcoming, accepted, awaiting publication	accepted, not yet printed	Two of the papers cited here are still in press and carry no page numbers.
A fixed phrase with no article, used after 'be' and inside citations in place of a year: 'Brown, in press'. It means the decision has been made, which is why it can be cited; a manuscript under review cannot.
abstract (of a paper)formal /ˈæbstrækt/ a short summary at the start of a paper giving its aim and results Synonyms: summary, synopsis, outline	a short summary at the start	The abstract promises more than the results section actually delivers.
The everyday adjective 'abstract', the opposite of concrete, is a different word: this noun is the summary at the front of a paper and takes first-syllable stress. Most journals cap it at 150 to 250 words, and a 'structured abstract' uses fixed headings such as Method and Results.
to outlineformal /ˈaʊtˌlaɪn/ describe the main points of something without the details Synonyms: sketch, summarize, set out	give the main points	Section two outlines the three positions taken in the debate so far.
The verb and noun are spelled and stressed alike, with the stress on 'out'. It promises the shape of an argument rather than the whole of it, so it suits an introduction. Compare 'detail', which promises the opposite.
corpusformal /ˈkɔrpəs/ a large collection of texts or records gathered for study Synonyms: body of texts, collection, archive	a collected body of texts	The corpus contains eight million words of spoken American English.
The plural is corpora, /ˈkɔrpərə/, and 'corpuses' is rare in linguistics. It usually names language material, though 'a corpus of letters' works for any body of documents. 'Corpus-based' describes work built on such a collection.
meta-analysisformal /ˌmɛtə əˈnæləsɪs/ a study that combines the results of many earlier studies statistically Synonyms: pooled analysis, combined study	a study combining studies	A meta-analysis of 42 trials puts the average benefit at about four percent.
The plural is meta-analyses, /əˈnæləsiz/, with the final syllable rhyming with 'seas'. American journals keep the hyphen. It combines results numerically, which is what separates it from a review that merely describes them.
systematic reviewformal /ˌsɪstəˈmætɪk rɪˈvju/ a review that searches for and judges all studies on a question by fixed rules Synonyms: structured review, comprehensive review	a rule-based survey of studies	The systematic review identified 311 papers and kept 27 for close reading.
The search rules are written before the search begins, which is what separates it from a review where the author simply chooses. It need not combine numbers; adding that step makes it a meta-analysis.
operational definitionformal /ˌɑpəˈreɪʃənəl ˌdɛfəˈnɪʃən/ a definition that says exactly how something will be measured in a study Synonyms: a working definition, measurable definition	a definition you can measure	The operational definition of poverty here is an income below half the national median.
It answers 'how will you know?' rather than 'what does it mean?', so two studies can share an idea and still disagree here. Writers introduce one with 'we operationalize trust as'. Attacking a paper's operational definition means attacking the fit between idea and measure.
parsimonyformal /ˈpɑrsəˌmoʊni/ the idea that the simplest explanation which fits is the best one Synonyms: economy of explanation, simplicity	using the fewest assumptions	Parsimony favors the shorter model, since the extra term explains almost nothing.
In everyday use it means stinginess with money, and the academic sense borrows only the thrift: fewer parts, not smaller ambitions. The adjective is parsimonious, as in 'the more parsimonious model'. It is a preference between explanations, never proof that one is right.
modelneutral /ˈmɑdəl/ a simplified description of how something works, used to explain or predict Synonyms: representation, simplified account, scheme	a simplified representation	The model fits the data well but assumes that costs stay constant.
Every model leaves things out on purpose, so 'simplified' is not a criticism of one. The standard verbs are 'fit', 'specify', 'estimate' and 'test a model', and it can be a statistical equation or a verbal account.
selection biasformal /sɪˈlɛkʃən ˈbaɪəs/ error caused by studying a group that is not typical of the whole Synonyms: selection effect, unrepresentative sampling	bias from who was included	Selection bias is the obvious worry: only the schools already improving agreed to take part.
It is about who ends up in the study, not about how they were measured, which separates it from measurement error. Volunteers, survivors and people who answer the phone are the classic sources. A design can carry it even when every number is recorded correctly.
to conceptualizeformal /kənˈsɛptʃuəˌlaɪz/ form a clear idea of something and say what it includes Synonyms: define, think of, frame	work out as an idea	Earlier work conceptualized trust as a single quality rather than several.
The frame is 'conceptualize X as Y', which pairs with 'operationalize X as Y': one settles what the idea is, the other how it will be measured. British spelling is 'conceptualise'; the noun is conceptualization.
taxonomyformal /tækˈsɑnəmi/ an ordered system for sorting things into groups Synonyms: classification, ordered system, scheme	an ordered classification	The authors propose a taxonomy of four error types, each with its own test.
It implies a worked-out structure with levels, not a simple list, which is why 'propose a taxonomy' is a substantial claim. Followed by 'of': 'a taxonomy of motives'. The adjective is taxonomic.
blindingformal /ˈblaɪndɪŋ/ keeping those involved from knowing which group a case is in Synonyms: masking, withholding group information	keeping people unaware	Blinding was impossible here, since everyone could see which classroom had the new books.
Single blinding hides the group from the participant, double blinding from the researcher as well, and a study that manages neither is open-label. Masking is now often preferred in medical writing. It guards against expectation rather than against chance.
synthesisformal /ˈsɪnθəsɪs/ the joining of separate findings or ideas into one account Synonyms: combination, integration, bringing together	a combining into a whole	What the field needs is a synthesis of these scattered results, not another small study.
The plural is syntheses, /ˈsɪnθəsiz/. Uncountable when it names the activity, countable when it names a particular account. Its opposite in argument is analysis, which breaks a whole into parts.
to synthesizeformal /ˈsɪnθəˌsaɪz/ bring separate findings together into one account Synonyms: combine, integrate, bring together	combine into one account	The review synthesizes results from twelve countries into a single picture.
Takes a direct object: 'synthesize the literature', 'synthesize findings across studies'. It asks for a new whole, not a list, which is why a summary is not a synthesis. A chemical sense means to make by combining substances.
cutoffneutral /ˈkʌtˌɔf/ the point on a scale at which cases are treated as one thing or the other Synonyms: threshold, dividing line	the dividing point	The cutoff for a pass was set at sixty percent before any of the papers were marked.
Choosing it after seeing the data is a well-known way to manufacture a result, so honest papers say when the cutoff was fixed. One word as a noun, two as the verb it came from. Often used before a noun: 'the cutoff score', 'the cutoff date'.

English C1 — Academic & Research Official

Rate this Deck

Word List — click grammar notes to expand

More English Decks

English C1

English C1 — Professional Communication

Business English — Negotiation & Leadership

Academic English — Research & Essays

English B2

English B2 — Business & Economy