Lecture 5 mcqs part 2

Question 1

1. Which of the following is true about Pearson’s correlation coefficient (r)?
   A) It reflects non-linear relationships
   B) It ranges from –2 to +2
   C) It is unitless and ranges between –1 and +1
   D) It shows the probability of an association

Answer 1

C

Question 2

42. A linear regression model finds:
    Tooth decay score = 3 + 1.5 × (sugar servings per day)
    How much higher would the decay score be for someone eating 4 servings/day vs. 1 serving/day?
    A) 1.5
    B) 4.5
    C) 6.0
    D) 3.0

Answer 2

D

Question 3

43. What does the intercept (constant) represent in a linear regression model?
    A) The outcome when all predictors = 0
    B) The slope of the regression line
    C) The strongest predictor
    D) The average value of the predictor

Answer 3

A

Question 4

44. A regression line fits a dataset with an r² of 0.82. How should you interpret this?
    A) 82% of the predictor variables are correlated
    B) 82% of the variation in Y is explained by X
    C) The correlation is 82%
    D) The error rate is 82%

Answer 4

B

Question 5

45. Which plot would you use to visually assess correlation?
    A) Bar chart
    B) Line graph
    C) Scatter plot
    D) Kaplan-Meier curve

Answer 5

C

Question 6

46. In a cohort study, the relative risk of tooth loss in smokers vs. non-smokers is 2.8. What does this imply?
    A) Smokers are 2.8 times more likely to keep their teeth
    B) The association is protective
    C) Smokers are 2.8 times more likely to lose their teeth
    D) The result is non-significant

Answer 6

C

Question 7

47. An odds ratio of 1.0 means:
    A) Strong positive association
    B) Strong negative association
    C) No association between exposure and outcome
    D) Statistically significant difference

Answer 7

C

Question 8

48. In survival analysis, if no events are recorded after 12 months, the Kaplan-Meier curve will:
    A) Rise steeply
    B) Flatten
    C) Drop sharply
    D) Reset to zero

Answer 8

B

Question 9

49. A study uses multivariable regression to adjust for age, gender, and diet. This helps to:
    A) Increase power
    B) Control for confounding
    C) Eliminate censoring
    D) Transform continuous variables

Answer 9

B

Question 10

50. If the confidence interval for an odds ratio is 1.2 to 3.6, what conclusion is most valid?
    A) The OR is not statistically significant
    B) There is likely a true association
    C) The odds are equal across groups
    D) The null hypothesis cannot be rejected

Answer 10

B

Question 11

51. Which type of analysis is most appropriate for:
    “Time until first recurrence of oral ulcers”?
    A) Logistic regression
    B) Relative risk
    C) Kaplan-Meier survival analysis
    D) Pearson’s correlation

Answer 11

C

Question 12

52. An odds ratio of 3.0 means:
    A) The odds of exposure are 3 times higher in controls
    B) The odds of the outcome are 3 times higher in the exposed group
    C) The risk is tripled
    D) 3 out of every 10 people have the condition

Answer 12

B

Question 13

53. What is the main limitation of Kaplan-Meier curves?
    A) They require normal distribution
    B) They cannot handle categorical outcomes
    C) They don’t adjust for other variables
    D) They can’t estimate survival probabilities

Answer 13

C

Question 14

54. You analyse gum disease and find:
    Odds ratio = 2.5, 95% CI = 0.95–4.2. What does this mean?
    A) There is a statistically significant association
    B) No clear conclusion can be made
    C) There is no association
    D) The result proves causation

Answer 14

B

Question 15

55. Which of the following best describes censoring in survival analysis?
    A) The study ends before an event occurs for a participant
    B) A participant experiences the event
    C) A participant has multiple events
    D) All participants are followed for the same time

Answer 15

A

Question 16

56. A dentist conducts a cross-sectional study on erosion prevalence. What measure is most appropriate?
    A) Relative risk
    B) Incidence rate
    C) Odds ratio
    D) Prevalence

Answer 16

D

Question 17

57. The hazard in survival analysis refers to:
    A) The rate at which people survive
    B) The cumulative chance of being censored
    C) The instantaneous risk of an event at a given time
    D) The difference in medians

Answer 17

C

Question 18

58. Which would increase the risk of a false conclusion in regression analysis?
    A) Checking residuals
    B) Ignoring non-linearity
    C) Including confounders
    D) Using a large sample

Answer 18

B

Question 19

59. When plotting a survival curve, a drop represents:
    A) A participant withdrawing
    B) A participant censored
    C) An event occurrence
    D) A new participant added

Answer 19

C

Question 20

60. What does an r² of 0.00 indicate in a regression model?
    A) The model perfectly fits the data
    B) All variability is explained
    C) The model explains none of the outcome variability
    D) The correlation is strong

Answer 20

C

Question 21

61. A correlation coefficient of –0.9 means:
    A) A weak inverse association
    B) A strong positive linear relationship
    C) A strong negative linear relationship
    D) No linear relationship

Answer 21

c

Question 22

62. Which value of r² would indicate the worst fit of a regression model?
    A) 0.72
    B) 0.00
    C) 0.45
    D) 0.10

Answer 22

B

Question 23

63. A study reports that for every additional cigarette smoked per day, bleeding score increases by 0.8 units. What is this 0.8 value?
    A) Correlation
    B) Intercept
    C) Regression coefficient
    D) Relative risk

Answer 23

C

Question 24

64. What does a scatter plot showing a tight downward sloping line suggest?
    A) No association
    B) Strong negative correlation
    C) High standard deviation
    D) Linear regression is inappropriate

Answer 24

B

Question 25

65. Which scenario would best justify using multiple regression instead of simple regression? A) Examining time to death B) Comparing two group means C) Predicting gum disease using age, diet, and brushing frequency D) Estimating prevalence

Answer 25

C

Question 26

66. An odds ratio of 0.6 with a 95% CI of 0.3–0.9 implies: A) Increased risk B) Protective effect, statistically significant C) No effect D) Strong positive associatioN

Answer 26

B

Question 27

67. What makes the odds ratio different from the relative risk? A) OR always underestimates risk B) RR is only used in case-control studies C) OR is appropriate for case-control studies; RR for cohort studies D) OR and RR are interchangeable

Answer 27

C

Question 28

68. A survival analysis shows the treatment group has a higher curve than the control group. This means: A) Worse survival in the treatment group B) Higher risk of the event C) Longer survival in the treatment group D) More censoring in the control group

Answer 28

C

Question 29

69. What kind of outcome is required to use Kaplan-Meier survival analysis? A) Ordinal B) Binary C) Time-to-event D) Continuous

Answer 29

C

Question 30

70. Which of the following is NOT a function of Cox proportional hazards regression? A) Adjusting for multiple covariates B) Estimating hazard ratios C) Measuring cumulative incidence D) Handling censored data

Answer 30

c

Question 31

71. In a survival plot, what does a vertical tick mark represent? A) An event B) A lost subject C) A censored subject D) End of follow-up

Answer 31

C

Question 32

72. Which of the following would likely result in a non-significant odds ratio? A) Wide confidence interval including 1 B) OR = 3.5, 95% CI: 2.1–5.0 C) Narrow CI, far from 1 D) p-value < 0.05

Answer 32

A

Question 33

73. In regression, which assumption is checked by plotting residuals vs. predicted values? A) Normality B) Homoscedasticity C) Multicollinearity D) Linearity of predictors

Answer 33

B

Question 34

74. In the equation: Bleeding index = 2 + 0.5 × brushing frequency – 0.3 × flossing frequency What effect does flossing have? A) No effect B) It increases bleeding index C) It decreases bleeding index D) Cannot be interpreted

Answer 34

C

Question 35

75. Which result suggests the strongest linear association? A) r = 0.89 B) r = –0.92 C) r = 0.05 D) r = 0.67

Answer 35

B

Question 36

76. Which situation is most appropriate for a hazard ratio? A) Estimating odds of caries B) Comparing time to infection between two dental treatments C) Assessing prevalence of periodontal disease D) Predicting bleeding scores using regression

Answer 36

B

Question 37

77. A p-value of 0.001 associated with an OR of 2.2 suggests: A) The OR is likely due to chance B) A statistically significant association C) The sample size is too small D) The result is not meaningful

Answer 37

B

Question 38

78. What type of variable is best suited for linear regression? A) Nominal B) Binary C) Continuous D) Ordinal

Answer 38

C

Question 39

79. A regression output gives β = –1.3, p = 0.70. What does this mean? A) Significant inverse relationship B) No statistically significant association C) Large sample size D) OR = –1.3

Answer 39

B

Question 40

80. When is Cox regression preferred over Kaplan-Meier? A) When you want to show time-to-event without adjustment B) When your outcome is binary C) When adjusting for multiple predictors of time-to-event D) When there is no censoring

Answer 40

C

Question 41

81. A study on oral health finds r² = 0.67 for sugar intake and decay. How should you interpret this? A) Sugar intake predicts 67% of decay variation B) 67% of sugar intake is explained by decay C) There is no correlation D) This is a weak association

Answer 41

A

Question 42

82. A logistic regression model is used to predict gum disease (yes/no) based on smoking and brushing habits. What type of outcome is being modelled? A) Continuous B) Binary C) Ordinal D) Nominal with >2 categories

Answer 42

B

Question 43

83. In regression, multicollinearity refers to: A) Missing data in the outcome B) A non-linear relationship between X and Y C) Predictors being highly correlated with each other D) Unequal variances of residuals

Answer 43

C

Question 44

84. In a Kaplan-Meier curve, the median survival time is: A) When 75% of participants remain event-free B) When half the events are censored C) When 50% of the sample has experienced the event D) The end of follow-up

Answer 44

C

Question 45

85. What does a wide confidence interval for a regression coefficient usually suggest? A) High precision B) Small effect size C) Greater uncertainty D) Low standard error

Answer 45

C

Question 46

86. A study uses Cox regression and finds HR = 0.65 for brushing ≥2x/day vs. <2x/day. What does this mean? A) Brushing more is harmful B) There's no effect C) Higher brushing frequency is associated with lower hazard D) Hazard is increased in frequent brushers

Answer 46

C

Question 47

87. Which regression method is best suited to predict a continuous outcome using several variables? A) Logistic regression B) Cox regression C) Linear regression D) Chi-square test

Answer 47

C

Question 48

88. A sample has an OR of 1.5 but a p-value of 0.52. What should be concluded? A) Strong association B) Statistically significant C) Likely due to chance D) The null hypothesis is rejected

Answer 48

C

Question 49

89. What is an advantage of using Cox regression over Kaplan-Meier curves? A) Simpler interpretation B) It allows adjusting for covariates C) No need for censoring D) Assumes linearity

Answer 49

B

Question 50

90. A survival curve that remains flat for several months indicates: A) Many events occurred early B) No events occurred during that time C) High mortality D) Frequent censoring

Answer 50

B