lecture 3 mcqs part 2 Flashcards
(40 cards)
1
Q
. What is the definition of a p-value?
A) The chance the alternative hypothesis is true
B) The probability of the null hypothesis being correct
C) The probability of the observed data or more extreme, assuming the null hypothesis is true
D) The probability of a Type II error
A
C
2
Q
- Which of the following best defines the null hypothesis (H₀)?
A) There is an effect in the population
B) The result in the sample is statistically significant
C) There is no real effect, and the observed effect is due to chance
D) We believe the observed difference reflects a real population difference
A
C
3
Q
- In the e-Learning example, what was the observed success rate for implant treatment?
A) 75%
B) 90%
C) 85%
D) 80%
A
C
4
Q
- Which scenario describes a Type I error?
A) Failing to detect an effect when one exists
B) Finding a statistically significant effect when none exists
C) Correctly rejecting the null hypothesis
D) Misinterpreting clinical significance
A
B
5
Q
- What is a key property of standard error?
A) It increases with sample size
B) It reflects the accuracy of measuring one individual
C) It is a measure of variability in a population
D) It reflects the variability of a statistic (e.g., sample mean or proportion) across samples
A
D
6
Q
- If we increase the sample size, what generally happens to the standard error?
A) It increases
B) It stays the same
C) It decreases
D) It fluctuates randomly
A
C
7
Q
- If the p-value is greater than 0.05, what conclusion is typically made?
A) We reject the null hypothesis
B) The result is clinically significant
C) There is no statistically significant evidence against the null hypothesis
D) We accept the null hypothesis as true
A
C
8
Q
- The research hypothesis (H₁) assumes:
A) The sample data has errors
B) The null hypothesis is correct
C) The observed effect reflects a real difference in the population
D) Sampling variability is minimized
A
C
9
Q
- What does a p-value of 0.003 mean in terms of statistical significance?
A) There’s a 3% chance the null is true
B) The result is statistically significant at the 0.05 level
C) The difference is likely due to random variation
D) The sample size is too small to detect an effect
A
B
10
Q
- Which of the following is true about statistical significance?
A) It proves a causal relationship
B) It always means clinical importance
C) It only occurs with large sample sizes
D) It shows the data are unlikely under the null hypothesis
A
D
11
Q
- What does “clinically significant” mean?
A) The result is statistically significant
B) The p-value is below 0.05
C) The observed effect is meaningful in practice
D) The study used computer simulation
A
C
12
Q
- What would a lack of statistical significance mean?
A) The null hypothesis is true
B) There is no effect in the population
C) There’s insufficient evidence to reject the null
D) The alternative hypothesis is proven false
A
C
13
Q
- Which value reflects standard error of difference in proportions?
A) Variance
B) Range
C) Sampling variability
D) 3.4% (2 standard errors)
A
D
14
Q
- In hypothesis testing, what does a simulation-based p-value represent?
A) How many successful outcomes were observed
B) The proportion of simulated results as extreme as the observed one
C) The difference in means between two groups
D) The maximum likelihood of the null hypothesis
A
B
15
Q
- In the e-Learning 3 example, the difference between implant and endodontic success rates was:
A) 10%
B) 15%
C) 5%
D) 20%
A
C
16
Q
- What is the term for false negatives in hypothesis testing?
A) Type I error
B) Power
C) Type II error
D) Confidence interval
A
C
17
Q
- If the study size is small, what tends to happen to the p-value?
A) It’s always significant
B) It tends to be larger due to increased variability
C) It is unaffected
D) It becomes more accurate
A
B
18
Q
- In the e-Learning simulation with 50 per group, the variability of the result:
A) Decreased
B) Stayed constant
C) Increased
D) Made no difference to significance
A
C
19
Q
- What is the correct interpretation of power in hypothesis testing?
A) Probability of detecting a true effect if one exists
B) Probability of making a Type I error
C) Likelihood that the null hypothesis is true
D) Chance of false positive
A
a
20
Q
- If a study has power of 80%, what is the chance of a Type II error?
A) 5%
B) 10%
C) 20%
D) 80%
A
C
21
Q
- Which of the following is not affected by sample size?
A) Power
B) Clinical significance
C) Standard error
D) p-value
A
B
22
Q
- What should always be done before checking p-values?
A) Examine the data
B) Run a t-test
C) Estimate standard deviation
D) Simulate under the null
A
A
23
Q
- What is the main goal of hypothesis testing?
A) Evaluate evidence against the null hypothesis
B) Describe the sample
C) Determine confidence intervals
D) Choose a study design
A
A
24
Q
- What does standard error represent?
A) Standard deviation of a statistic across repeated samples
B) Error in individual measurement
C) The population mean
D) Type I error rate
A
A
25
25. Which is a feature of the null hypothesis?
A) It assumes no effect in the population
B) It is what the researcher hopes to prove
C) It predicts a specific difference
D) It describes the sample
A
26
26. If a simulation shows 4 out of 1000 results as extreme as the observed result, what is the p-value?
A) 0.004
B) 0.01
C) 0.04
D) 0.40
A
27
27. What is the appropriate conclusion when p < 0.05?
A) Reject the null hypothesis
B) The sample is biased
C) Accept the null hypothesis
D) The result is inconclusive
a
28
28. Which factor does not directly influence the p-value?
A) Sample size
B) Effect size
C) Power
D) Standard error
C
29
9. A statistically significant result that has no clinical meaning is described as:
A) Overpowered
B) Statistically but not clinically significant
C) False positive
D) Clinically invalid
B
30
30. A Type II error occurs when:
A) A false null is rejected
B) A p-value is below 0.01
C) A false null is not rejected
D) The null hypothesis is true
C
31
31. What does “power” represent in statistical testing?
A) The ability to detect an effect when one truly exists
B) The likelihood of rejecting a true null
C) The effect size of the outcome
D) The amount of error in the data
A
32
32. When is a two-tailed test appropriate?
A) When the direction of the difference is predicted
B) When the null is false
C) When testing for differences in either direction
D) When only one group is studied
C
33
33. What does a p-value of 0.10 indicate (with α = 0.05)?
A) The test is statistically significant
B) There is not enough evidence to reject the null hypothesis
C) The sample size is too large
D) The null hypothesis is proven true
B
34
34. If simulations show 5 in 1000 results as extreme as the observed, what’s the p-value?
A) 0.005
B) 0.050
C) 0.500
D) 0.0500
A
35
35. Which of the following can reduce the risk of a Type II error?
A) Lowering α to 0.01
B) Increasing sample size
C) Using a one-tailed test
D) Raising standard error
B
36
36. A small difference between groups is found, but it’s within expected sampling variability. What does this suggest?
A) The study is flawed
B) Evidence against the null is strong
C) No strong evidence to reject the null
D) The result is statistically significant
C
37
37. Why does a smaller sample size result in greater variability in simulation-based inference?
A) Fewer randomisation attempts
B) Larger standard error
C) Higher p-values
D) Reduced Type I error
B
38
38. What is the Type I error rate if α = 0.05?
A) 1%
B) 50%
C) 5%
D) 95%
C
39
39. What is the main function of hypothesis testing?
A) Estimate confidence intervals
B) Measure standard deviation
C) Evaluate evidence against the null hypothesis
D) Create a research question
C
40
40. A 5% difference is observed between two groups. What must be known to determine if it’s significant?
A) Confidence interval
B) P-value
C) Sample size
D) Clinical importance
B
