Data, research Interpretation and Evidence Based Dentistry Quiz 2

Question 1 of 40

1. Question

Null hypothesis meaning

A study compares mean probing depth between two scaling techniques. The null hypothesis (H₀) usually states that:

Technique A is superior to Technique B

Correct Incorrect Correct answer

Technique B is superior to Technique A

Correct Incorrect Correct answer

There is no true difference in mean probing depth between techniques

Correct Incorrect Correct answer

Both techniques are ineffective

Correct Incorrect Correct answer

Question 2 of 40

2. Question

Type I error

In a randomized trial comparing two mouthrinses, researchers conclude there is a difference, but in reality there is none. This is:

Type I error (α error)

Correct Incorrect Correct answer

Type II error (β error)

Correct Incorrect Correct answer

Correct acceptance of H₀

Correct Incorrect Correct answer

Lack of power

Correct Incorrect Correct answer

Question 3 of 40

3. Question

Type II error

A small trial finds no significant difference between two caries-prevention programs, but a true difference does exist. This is:

Type I error

Correct Incorrect Correct answer

Type II error

Correct Incorrect Correct answer

High power

Correct Incorrect Correct answer

Confounding

Correct Incorrect Correct answer

Question 4 of 40

4. Question

p-value interpretation

A periodontal trial reports a p-value of 0.02 when comparing bleeding indices between two treatments (α = 0.05). Which interpretation is most appropriate?

There is a 2% chance that the alternative hypothesis is true

Correct Incorrect Correct answer

If the null hypothesis is true, there is a 2% chance of seeing results this extreme

Correct Incorrect Correct answer

The null hypothesis is definitely false

Correct Incorrect Correct answer

The study is biased

Correct Incorrect Correct answer

Question 5 of 40

5. Question

Confidence interval and 0

A study reports that a new desensitizing paste decreases hypersensitivity by 1.5 units (95% CI: 0.2 to 2.8 units) compared with control. At α = 0.05, this result is:

Not significant because CI is wide

Correct Incorrect Correct answer

Statistically significant because CI does not include 0

Correct Incorrect Correct answer

Not significant because lower bound is small

Correct Incorrect Correct answer

Inconclusive

Correct Incorrect Correct answer

Question 6 of 40

6. Question

Which change will usually increase the power of a clinical trial?

Reducing sample size

Correct Incorrect Correct answer

Increasing the variability of measurements

Correct Incorrect Correct answer

Using a smaller effect size to detect

Correct Incorrect Correct answer

Increasing sample size

Correct Incorrect Correct answer

Question 7 of 40

7. Question

Prevalence vs incidence

A cross-sectional survey finds that 20% of adults in a community currently have untreated caries. This is a measure of:

Incidence

Correct Incorrect Correct answer

Prevalence

Correct Incorrect Correct answer

Relative risk

Correct Incorrect Correct answer

Odds ratio

Correct Incorrect Correct answer

Question 8 of 40

8. Question

Risk ratio

In a cohort study, 30 of 150 smokers and 10 of 150 non-smokers develop periodontitis. The risk ratio (RR) is:

(30/150) ÷ (10/150) = 3.0

Correct Incorrect Correct answer

(10/150) ÷ (30/150) = 0.33

Correct Incorrect Correct answer

(30/10) = 3.0 but irrelevant

Correct Incorrect Correct answer

(150/30) = 5.0

Correct Incorrect Correct answer

Question 9 of 40

9. Question

Odds ratio in case–control

In a case–control study of oral cancer, 40 of 60 cases and 10 of 60 controls are heavy alcohol users. The odds ratio (OR) for heavy alcohol use is:

4

Correct Incorrect Correct answer

2

Correct Incorrect Correct answer

1

Correct Incorrect Correct answer

6

Correct Incorrect Correct answer

Question 10 of 40

10. Question

Sensitivity definition

Sensitivity of a diagnostic test is:

P(test positive | disease present)

Correct Incorrect Correct answer

P(test negative | disease present)

Correct Incorrect Correct answer

P(disease present | test positive)

Correct Incorrect Correct answer

P(disease absent | test negative)

Correct Incorrect Correct answer

Question 11 of 40

11. Question

Specificity definition

Specificity is:

P(test positive | no disease)

Correct Incorrect Correct answer

P(test negative | no disease)

Correct Incorrect Correct answer

P(disease | test negative)

Correct Incorrect Correct answer

P(disease | test positive)

Correct Incorrect Correct answer

Question 12 of 40

12. Question

Predictive values & prevalence

When a disease becomes more prevalent in the population, what typically happens?

Positive predictive value increases

Correct Incorrect Correct answer

Negative predictive value increases

Correct Incorrect Correct answer

Both remain unchanged

Correct Incorrect Correct answer

Test sensitivity decreases

Correct Incorrect Correct answer

Question 13 of 40

13. Question

Normal distribution

In a perfectly normal distribution, which is true?

Mean > median > mode

Correct Incorrect Correct answer

Mean < median < mode

Correct Incorrect Correct answer

Mean = median = mode

Correct Incorrect Correct answer

Only median = mode

Correct Incorrect Correct answer

Question 14 of 40

14. Question

Outliers and median

A DMFT dataset has a few extremely high values. Which is least affected by these outliers?

Mean

Correct Incorrect Correct answer

Standard deviation

Correct Incorrect Correct answer

Median

Correct Incorrect Correct answer

Range

Correct Incorrect Correct answer

Question 15 of 40

15. Question

Interquartile range

The interquartile range (IQR) is:

Maximum minus minimum

Correct Incorrect Correct answer

Mean plus/minus SD

Correct Incorrect Correct answer

75th percentile minus 25th percentile

Correct Incorrect Correct answer

Median of all observations

Correct Incorrect Correct answer

Question 16 of 40

16. Question

Correlation coefficient

A study finds r = 0.8 between time spent studying and exam score. Which statement best describes this?

Strong positive linear relationship

Correct Incorrect Correct answer

Strong negative linear relationship

Correct Incorrect Correct answer

No correlation

Correct Incorrect Correct answer

Perfect positive correlation

Correct Incorrect Correct answer

Question 17 of 40

17. Question

r² meaning

If r = 0.6 between plaque index and bleeding score, what does r² = 0.36 mean?

36% of variation in bleeding is explained by plaque levels

Correct Incorrect Correct answer

60% of variation in plaque is explained by bleeding

Correct Incorrect Correct answer

36% of patients have bleeding

Correct Incorrect Correct answer

No useful information

Correct Incorrect Correct answer

Question 18 of 40

18. Question

Choosing a test – two independent means

To compare mean pocket depth between treatment and control groups (different patients, approximately normal data), the best test is:

Paired t-test

Correct Incorrect Correct answer

Independent samples t-test

Correct Incorrect Correct answer

Chi-square test

Correct Incorrect Correct answer

Wilcoxon signed-rank test

Correct Incorrect Correct answer

Question 19 of 40

19. Question

Chi-square test usage

Chi-square test is most appropriate for:

Comparing mean DMFT between groups

Correct Incorrect Correct answer

Testing association between smoking (yes/no) and periodontitis (yes/no)

Correct Incorrect Correct answer

Calculating correlation between two continuous variables

Correct Incorrect Correct answer

Estimating survival curves

Correct Incorrect Correct answer

Question 20 of 40

20. Question

Paired data example

Which scenario calls for a paired t-test?

Comparing mean DMFT in two different schools

Correct Incorrect Correct answer

Comparing plaque scores before and after a mouthrinse in the same patients

Correct Incorrect Correct answer

Comparing pocket depths in smokers vs non-smokers

Correct Incorrect Correct answer

Comparing 3 materials at one time point in different patients

Correct Incorrect Correct answer

Question 21 of 40

21. Question

ANOVA purpose

A researcher wants to compare mean pain scores for three different anesthetic techniques. Which test should they use?

Paired t-test

Correct Incorrect Correct answer

One-way ANOVA

Correct Incorrect Correct answer

Chi-square

Correct Incorrect Correct answer

Mann–Whitney U

Correct Incorrect Correct answer

Question 22 of 40

22. Question

Cross-sectional study strength

Cross-sectional studies are particularly useful for:

Establishing temporal cause-and-effect

Correct Incorrect Correct answer

Measuring prevalence of disease and associated factors at one time point

Correct Incorrect Correct answer

Evaluating incidence over many years

Correct Incorrect Correct answer

Minimizing confounding

Correct Incorrect Correct answer

Question 23 of 40

23. Question

Cohort vs case–control

A cohort study typically:

cohort study typically: A. Starts with diseased and non-diseased people and looks back in time

Correct Incorrect Correct answer

Randomly allocates treatment

Correct Incorrect Correct answer

Follows exposed and unexposed groups forward to see who develops disease

Correct Incorrect Correct answer

Samples only at a single time point

Correct Incorrect Correct answer

Question 24 of 40

24. Question

Recall bias

In a study of oral cancer, patients with cancer remember their alcohol intake more intensely than controls. This is:

Selection bias

Correct Incorrect Correct answer

Recall bias

Correct Incorrect Correct answer

Measurement bias

Correct Incorrect Correct answer

Confounding

Correct Incorrect Correct answer

Question 25 of 40

25. Question

Confounding

A study shows that people who chew sugar-free gum have lower caries, but they also tend to brush more often. Toothbrushing frequency is:

Outcome

Correct Incorrect Correct answer

Confounder

Correct Incorrect Correct answer

Effect modifier only

Correct Incorrect Correct answer

Random error

Correct Incorrect Correct answer

Question 26 of 40

26. Question

Intention-to-treat

In a drug trial, some patients assigned to test drug switch to control mid-study. Intention-to-treat analysis means they are analyzed:

According to the drug actually taken

Correct Incorrect Correct answer

Only if they completed the study

Correct Incorrect Correct answer

In their originally assigned groups regardless of switches

Correct Incorrect Correct answer

Only after excluding noncompliant patients

Correct Incorrect Correct answer

Question 27 of 40

27. Question

NNT (number needed to treat)

A fluoride program reduces caries incidence from 40% to 20%. The NNT is:

1 / (0.40 – 0.20) = 5

Correct Incorrect Correct answer

1 / (0.20 – 0.40) = –5

Correct Incorrect Correct answer

1 / 0.40 = 2.5

Correct Incorrect Correct answer

1 / 0.20 = 5

Correct Incorrect Correct answer

Question 28 of 40

28. Question

Forest plot – CI crossing 1

In a meta-analysis, a relative risk with 95% CI 0.8–1.3 is shown; the CI crosses 1.0. Interpretation?

There is a statistically significant effect

Correct Incorrect Correct answer

No statistically significant effect at α = 0.05

Correct Incorrect Correct answer

The study is invalid

Correct Incorrect Correct answer

The drug is definitely harmful

Correct Incorrect Correct answer

Question 29 of 40

29. Question

Sample size & CI width

If a researcher doubles the sample size while variability remains constant, the confidence interval will:

Widen

Correct Incorrect Correct answer

Narrow

Correct Incorrect Correct answer

Stay the same

Correct Incorrect Correct answer

Disappear

Correct Incorrect Correct answer

Question 30 of 40

30. Question

Descriptive vs inferential statistics

Which is descriptive, not inferential?

95% confidence interval for mean plaque score

Correct Incorrect Correct answer

p-value for difference in means

Correct Incorrect Correct answer

Mean and standard deviation of DMFT in the sample

Correct Incorrect Correct answer

Regression coefficient for a predictor

Correct Incorrect Correct answer

Question 31 of 40

31. Question

Histogram vs bar chart

A histogram is most appropriate for:

Categorical data such as gender

Correct Incorrect Correct answer

Continuous data such as plaque scores grouped into intervals

Correct Incorrect Correct answer

Single proportions only

Correct Incorrect Correct answer

Paired nominal data

Correct Incorrect Correct answer

Question 32 of 40

32. Question

Blinding

In a double-blind RCT:

Only the patient is unaware of treatment

Correct Incorrect Correct answer

Only the investigator is unaware

Correct Incorrect Correct answer

Both patients and outcome assessors are unaware of group allocations

Correct Incorrect Correct answer

Everyone including manufacturers is always unaware

Correct Incorrect Correct answer

Question 33 of 40

33. Question

Survival analysis – censoring

In an implant survival study, a patient is lost to follow-up without failure at year 3. How is this handled?

Counted as failure at year 3

Correct Incorrect Correct answer

Excluded entirely from analysis

Correct Incorrect Correct answer

Censored at year 3

Correct Incorrect Correct answer

Treated as if followed for 10 years

Correct Incorrect Correct answer

Question 34 of 40

34. Question

Multiple comparisons problem

A researcher tests 30 outcomes at α = 0.05 without adjustment. Main concern?

Elevated risk of at least one Type I error

Correct Incorrect Correct answer

No effect on error rates

Correct Incorrect Correct answer

Only Type II errors increase

Correct Incorrect Correct answer

Only measurement error changes

Correct Incorrect Correct answer

Question 35 of 40

35. Question

Evidence levels

Which design generally provides the highest level of evidence for treatment efficacy?

Single case report

Correct Incorrect Correct answer

Cross-sectional study

Correct Incorrect Correct answer

Randomized controlled trial (or meta-analysis of RCTs)

Correct Incorrect Correct answer

Expert opinion

Correct Incorrect Correct answer

Question 36 of 40

36. Question

Systematic vs narrative review

A systematic review is characterized by:

Unstructured expert commentary

Correct Incorrect Correct answer

Explicit, pre-defined methods for searching, selecting, and appraising studies

Correct Incorrect Correct answer

Exclusion of all RCTs

Correct Incorrect Correct answer

Only anecdotal evidence

Correct Incorrect Correct answer

Question 37 of 40

37. Question

Bias vs random error

Bias in a study is best described as:

Random variation around the true value

Correct Incorrect Correct answer

Systematic deviation from the truth due to study design or conduct

Correct Incorrect Correct answer

Natural biological variability

Correct Incorrect Correct answer

Chance errors that decrease with increasing sample size

Correct Incorrect Correct answer

Question 38 of 40

38. Question

Clinical vs statistical significance

A very large trial finds a new rinse improves plaque index by 0.1 units vs control with p < 0.001. Which is most accurate?

The effect is statistically significant but may not be clinically meaningful

Correct Incorrect Correct answer

The effect is clinically huge because p is very small

Correct Incorrect Correct answer

The treatment is clearly superior for every patient

Correct Incorrect Correct answer

The result is not statistically significant

Correct Incorrect Correct answer

Question 39 of 40

39. Question

Evidence-based dentistry triad

Evidence-based dental decisions should integrate:

Only highest-quality research evidence

Correct Incorrect Correct answer

Clinical expertise and patient preferences only

Correct Incorrect Correct answer

Best available evidence, clinician experience, and patient values

Correct Incorrect Correct answer

Insurance coverage alone

Correct Incorrect Correct answer

Question 40 of 40

40. Question

Data interpretation – relative vs absolute effect

A mouthrinse reduces gingivitis prevalence from 10% to 5%. Which statement correctly distinguishes relative vs absolute risk reduction?

Absolute reduction = 50%, relative reduction = 5%

Correct Incorrect Correct answer

Absolute reduction = 5%, relative reduction = 50%

Correct Incorrect Correct answer

Both absolute and relative reduction are 5%

Correct Incorrect Correct answer

Both absolute and relative reduction are 50%

Correct Incorrect Correct answer

Data, research Interpretation and Evidence Based Dentistry Quiz 2

Quiz Summary

Information

Results

Results

Categories

1. Question

Null hypothesis meaning

2. Question

Type I error

3. Question

Type II error

4. Question

p-value interpretation

5. Question

Confidence interval and 0

6. Question

7. Question

Prevalence vs incidence

8. Question

Risk ratio

9. Question

Odds ratio in case–control

10. Question

Sensitivity definition

11. Question

Specificity definition

12. Question

Predictive values & prevalence

13. Question

Normal distribution

14. Question

Outliers and median

15. Question

Interquartile range

16. Question

Correlation coefficient

17. Question

r² meaning

18. Question

Choosing a test – two independent means

19. Question

Chi-square test usage

20. Question

Paired data example

21. Question

ANOVA purpose

22. Question

Cross-sectional study strength

23. Question

Cohort vs case–control

24. Question

Recall bias

25. Question

Confounding

26. Question

Intention-to-treat

27. Question

NNT (number needed to treat)

28. Question

Forest plot – CI crossing 1

29. Question

Sample size & CI width

30. Question

Descriptive vs inferential statistics

31. Question

Histogram vs bar chart

32. Question

Blinding

33. Question

Survival analysis – censoring

34. Question

Multiple comparisons problem

35. Question

Evidence levels

36. Question

Systematic vs narrative review

37. Question

Bias vs random error

38. Question