# ADAT practice questions: Data, research and evidence-based dentistry Ten original ADAT practice questions on data, research and evidence-based dentistry, each answered on this page with a rationale and a source. Last updated: 2026-09-17. ## Question 1 A learner asks whether an illustrated explanation helps adults perform a synthetic sorting task. The population, explanation and correct-completion outcome are stated. Which revision most directly supplies the missing comparison and time horizon? - A. Compare adults receiving the explanation with children receiving it, using each person’s most recent score. - B. Compare the explanation with written instructions alone, measuring correct completion four weeks after assignment. - C. Compare the explanation’s colors with its typography, measuring which feature participants prefer immediately. - D. Compare only adults who completed the explanation with one another, using their scores from any available visit. **Answer B:** Compare the explanation with written instructions alone, measuring correct completion four weeks after assignment. The question needs an alternative to the explanation and a defined assessment time. Written instructions alone and four weeks make the intended outcome contrast explicit without changing the stated population or outcome. Why the other choices do not fit: A: This changes the population comparison and leaves a common assessment time undefined; it does not identify the intended alternative explanation format. C: This changes the intervention contrast and substitutes preference for correct completion, answering a different question. D: This supplies no alternative program and lets follow-up vary. Completion-based selection also changes the group being compared. Source: [Cochrane Handbook chapter 3: Defining the criteria for including studies](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-03) ## Question 2 The question is why learners find an appointment-information sheet difficult to use at home. Which proposed source most directly addresses their experience and the context of that difficulty? - A. A randomized study reporting only the mean time needed to read two sheet formats. - B. A cohort reporting the proportion of learners who returned the sheet within a week. - C. A diagnostic study comparing a reading test with a specified reference assessment. - D. An interview study analyzing varied learners’ accounts of using the sheet at home. **Answer D:** An interview study analyzing varied learners’ accounts of using the sheet at home. The question asks about experience and context. Interviews with a transparent, question-appropriate analysis can explain what learners found difficult; the design still requires appraisal. Why the other choices do not fit: A: Reading time measures a quantitative contrast. Without accounts or another suitable inquiry, it does not explain the experience being asked about. B: Return frequency addresses behavior, not why the sheet was difficult to use. A cohort label does not make it the best fit for every question. C: Diagnostic correspondence answers a different question. It does not directly supply the requested accounts of use at home. Source: [Cochrane Handbook chapter 21: Qualitative evidence](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-21) ## Question 3 In a toy database, AND requires both term sets and OR includes either. The intended search requires either “coaching” or “instruction,” together with “dental.” Which expression preserves that grouping? - A. (coaching OR instruction) AND dental - B. (coaching AND instruction) AND dental - C. (coaching OR instruction) OR dental - D. (coaching AND dental) AND instruction **Answer A:** (coaching OR instruction) AND dental The OR group accepts either synonym, and AND then requires the dental concept. Parentheses make the intended combination explicit. Why the other choices do not fit: B: This requires both synonyms, excluding records that use only one of them. C: This also retrieves records with dental alone or either instructional term alone, without requiring both concepts. D: This again requires both coaching and instruction; relocating the parentheses does not create an alternative between them. Source: [NLM PubMed User Guide](https://pubmed.ncbi.nlm.nih.gov/help/) ## Question 4 A completely specified fictional benchmark has 80 unique records, V001–V080. Exactly V001–V016 are relevant. A search returns V001–V012 and V017–V044. What are its recall and precision, respectively? - A. 30% recall and 75% precision. - B. 15% recall and 30% precision. - C. 75% recall and 30% precision. - D. 75% recall and 50% precision. **Answer C:** 75% recall and 30% precision. There are 16 relevant records in the benchmark, 12 retrieved relevant records, and 12 + 28 = 40 retrieved records. Recall is 12/16 = 75%; precision is 12/40 = 30%. Why the other choices do not fit: A: This reverses the two denominators: 30% uses all retrieved records, whereas 75% uses all relevant records. B: 12/80 = 15% divides by the complete database rather than the relevant set. That is not recall. D: The recall is correct, but 50% is 40/80, the fraction of all records retrieved. Precision counts relevant records within the retrieved set. Source: [Cochrane Handbook chapter 4: Searching for and selecting studies](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-04) ## Question 5 A review prespecifies adults, format A versus B, and a four-week task outcome. An abstract describes adults and A versus B but gives neither outcome details nor follow-up time. What is the most defensible next step? - A. Exclude the study because the abstract proves that no four-week outcome was measured. - B. Obtain the full report or clarification before deciding whether the missing details satisfy eligibility. - C. Include the study in the four-week synthesis by treating the unreported result as no difference. - D. Replace the four-week criterion with any follow-up time so this abstract can be included. **Answer B:** Obtain the full report or clarification before deciding whether the missing details satisfy eligibility. The abstract lacks information needed to apply the rule. Further retrieval or clarification can distinguish absent reporting from incompatible methods; neither eligibility nor a numerical result can be inferred from silence. Why the other choices do not fit: A: An omitted detail is not proof that the outcome was never measured. C: Missing reporting supplies neither a four-week estimate nor a zero effect. It cannot be entered as an observed no-difference result. D: Changing the criterion simply to accommodate a report undermines the prespecified question and is unnecessary before inspecting the missing information. Source: [Cochrane Handbook chapter 3: Defining the criteria for including studies](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-03) ## Question 6 A secondary summary cites a 2022 methods document. The issuing organization’s page links that document to a 2025 correction, but the summary does not mention the correction. Before using the claim, what should the reader do? - A. Use the 2022 wording because the original publication has priority over every later notice. - B. Use the summary because its citation establishes that its author inspected all later versions. - C. Discard both documents because a correction establishes that all of the original work is unusable. - D. Inspect the original passage and correction, determining whether the correction changes the claim. **Answer D:** Inspect the original passage and correction, determining whether the correction changes the claim. Identity, version relationships and the actual passage matter. A correction can affect the proposed claim, a different passage or a limited detail; its scope must be read. Why the other choices do not fit: A: Original publication does not override a valid correction to that publication. B: A citation is not evidence that later notices or the supporting passage were actually checked. C: A correction is not automatically a complete invalidation. Discarding every claim without reading its scope is also unsupported. Source: [NLM PubMed User Guide](https://pubmed.ncbi.nlm.nih.gov/help/) ## Question 7 Two reports have the same registration identifier, recruitment dates and participants. One gives the methods and the other gives later outcomes. How should a review initially handle them? - A. Link both reports to one study and use their complementary information without treating them as independent replications. - B. Count two independent studies because the methods and outcomes were published in separate documents. - C. Discard the later report as a duplicate even if it contains outcomes missing from the earlier document. - D. Keep only the longer report because document length determines which participant information is authoritative. **Answer A:** Link both reports to one study and use their complementary information without treating them as independent replications. Matching study identifiers and participants indicate related reports. Linking preserves useful information while preventing duplicate counting of the same study or participants. Why the other choices do not fit: B: Documents and studies are different units. Separate publication does not make the underlying participants independent. C: Related reports can provide additional outcomes. Record deduplication is not a reason to discard relevant later information. D: Length does not establish which report contains the needed or corrected information; inspect and reconcile both. Source: [Cochrane Handbook chapter 4: Searching for and selecting studies](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-04) ## Question 8 A researcher broadens a search after noticing that a known relevant report uses a synonym absent from the first query. Which record best supports reproducibility of this revision? - A. The final count alone, because equal counts establish that two searches retrieved identical records. - B. The final query alone, because the original query is superseded and cannot explain selection decisions. - C. Both exact queries with database/platform, dates, limits, counts and the reason for adding the synonym. - D. The titles of the first few results, because top-ranked results establish the search’s overall completeness. **Answer C:** Both exact queries with database/platform, dates, limits, counts and the reason for adding the synonym. Saving both versions and their execution details makes the change inspectable. The reason explains the revision without pretending that a count establishes completeness. Why the other choices do not fit: A: Different result sets can have the same count. Counts do not encode the query or selection history. B: The original query explains how retrieval changed and should remain part of the audit trail. D: Search rank and a short title list neither reconstruct the query nor show that all relevant records were found. Source: [Cochrane Handbook chapter 4: Searching for and selecting studies](https://www.cochrane.org/authors/handbooks-and-manuals/handbook/current/chapter-04) ## Question 9 Investigators use archived enrollment records to identify everyone who entered programs A or B in 2021, then obtain each person’s later task outcome from 2022 records. Enrollment was not assigned by the investigators. Which design description fits? - A. A case-control study, because all the data already existed when the investigators began. - B. A randomized trial, because the investigators can compare two named programs. - C. A retrospective observational cohort, because a past program-entry cohort is followed through later records. - D. A cross-sectional survey, because program and outcome records are read on the same day. **Answer C:** A retrospective observational cohort, because a past program-entry cohort is followed through later records. The population is defined by past program entry and later outcomes are reconstructed. The timing of record review does not change this cohort direction, and the programs were not assigned by the investigators. Why the other choices do not fit: A: Case-control selection begins with outcome-defined samples. Existing records alone do not make a study case-control. B: Two comparison groups do not establish experimental assignment, much less random assignment. D: The day the analyst reads records is not the time structure of the study. The records encode entry followed by later outcome. Source: [CDC Field Epidemiology Manual: design and analysis chapters](https://www.cdc.gov/field-epi-manual/php/chapters/design-conduct-analyze-field-studies.html) ## Question 10 An investigator selects 40 people with task failure and 80 without task failure, then asks about previous program use. No population sampling fractions are supplied. What does 40/120 describe? - A. The case fraction in the selected sample, not an estimate of the source population’s failure risk. - B. The source population’s failure risk, because every selected participant appears in the denominator. - C. The failure risk among program users, because cases are selected before exposure is measured. - D. The incidence rate of failure, because the selected sample includes both cases and controls. **Answer A:** The case fraction in the selected sample, not an estimate of the source population’s failure risk. 40/120 = one third is the fraction of sampled people selected as cases. Outcome-based sampling and absent population sampling information prevent interpreting it as population risk. Why the other choices do not fit: B: Including all selected participants does not turn an outcome-selected sample into the source population’s at-risk denominator. C: The numerator and denominator are not restricted to program users; exposure has not even been tabulated. D: No person-time denominator is supplied. A sampled case fraction is not an incidence rate. Source: [CDC Field Epidemiology Manual: design and analysis chapters](https://www.cdc.gov/field-epi-manual/php/chapters/design-conduct-analyze-field-studies.html) ## Next step [Take the free 25-question ADAT practice test](https://dentovio.com/adat/free-practice-test) Official reference: [ADA 2026 ADAT Candidate Guide](https://www.ada.org/-/media/project/ada-organization/ada/ada-org/files/education/2026_adat_candidate_guide.pdf). Original exam-style questions written for study, never recalled exam content. Independent educational preparation, not affiliated with or endorsed by the American Dental Association. Practice results are not calibrated ADAT scores.