1. Science
  2. Other
  3. question 1 recall bias is a difference in the recall...

Question: question 1 recall bias is a difference in the recall...

Question details

Question 1

Recall Bias is:

   

a)     Difference in the recall of past exposures between those who have disease and those who are disease free.

   

b)     Covering up the fact that the study findings did not have any practical significance.

   

c)     Bad memory or being forgetful.

   

d)     When study subjects know the study hypothesis and they try to sabotage study data. 

   

e)     Under reporting of some exposures and past behaviours.

   

f)      Change in the participants' behavior when they know they are being observed.

Question 2

According to APA style a p = .000 in SPSS output is written as:

 

   

a)     <.01

   

b)     >.001

   

c)     <.001

   

d)     <.05

Question 3

A team of chocoholic researchers conducted a study where they recorded blood pressure of all the participants at baseline and then advised them to eat a certain amount and a certain brand of chocolate every day (along with a special diet) for 3 months. They recorded blood pressure at the end of three months, once again. Now they want to know if eating dark chocolate made any difference to the participants' blood pressure but they  need your help to identify which one of the following will be a suitable test. Please identify a test which is suitable for the purpose of this study.

   

a)     Pearson's Chi squared test

   

b)     Levene's test

   

c)     One sample t test

   

d)     Pearson's Correlation test

   

e)     Paired Samples t test

   

f)      Independent Samples t test

 

Question 4

Researchers reported that their randomized controlled trial clearly showed that the intervention to reduce cholesterol was effective. In order to make sure that the experimental and control groups were similar at baseline in terms of age, years of education, self-rated health score, anxiety scores and cholesterol levels they presented the results of these baseline comparisons in a table and provided p values (all >.05). Which one of the following tests listed below was used to compare the two groups at baseline?

 

   

a)     Independent Samples t test

   

b)     Dependent/Paired Samples t test

   

c)     Chi Squared test

   

d)     Pearson's Correlation coefficient

   

e)     Levene's test

Question 5

What will be a suitable test to compare the average mark for first written assignment with average mark over the previous five semesters?

   

a)     Pearson's Chi Square test

   

b)     Independent samples t test

   

c)     Paired samples t test

   

a)     One sample t test

Question 6

Results of a Paired Samples t test are statistically significant if:

   

p = <.05 and 95% CI for mean difference includes zero.

   

p = >.05 and 95% CI for mean difference does not include zero.

   

p = >.05 and 95% CI for mean difference includes zero.

   

p = <.05 and 95% CI for mean difference does not include zero.

Question 7

 

randomized controlled trial was conducted to find out if using two attempts for a series of ten quizzes achieved better results than using single attempt. A large cohort of students were randomized into two groups; students in group A (exposed) were told to use both attempts while students in group B (not exposed) had only one attempt per quiz (in case you are wondering these students were more than happy to just do all quizzes only once and have extra time for things they considered important). An overall grade of greater than 70% was considered the desirable 'outcome'. There were 1,192 students in group A and out of these 892 achieved greater than 70%, while there were a total of 1,192 students in group B and 110 scored higher than 70%. Calculate a suitable measure of association/effect to find out if those who used both attempts were more likely to achieve >70% compared to those who used single attempt & results are probably going to speak for themselves :). Only final answer is needed with two decimals after correct rounding.

 

 

=

Question 8

Not just the two sets of scores but their difference should also show roughly normal distribution. This assumption belongs to which one of the following tests:

   

 

Paired Samples t test

   

One Sample t test

   

Pearson's Chi Squared test

   

Independent Samples t test

   

Pearson's Correlation test

Question 9

It is important that we rule out three possible explanations before we can conclude that the observed effect (or the difference between intervention and control group) at the end of an experimental study is only due to the independent variable or the intervention. Please choose one option that has all three correct possible explanations. 

   

Random chance, Bias and Confounding.

   

Bias, Confounding and Systematic error.

   

Random chance, Random error and Confounding.

   

Random chance, Random error and Systematic error.

Question 10

Which one of the following test you will use to assess whether a characteristic shows normal distribution in the source population that is represented by the sample?

   

a)     Shapiro's Homoscedasticity test

   

b)     Chi Square test

   

c)     Shapiro-Wilk test

   

d)     Levene's test

   

e)     Pearson's Normality test

   

f)      Paired Samples t test

 

 

Question 11

When reporting the results and statistical significance of a Paired Samples t test what are the seven things that should be included in the description? (Heads up for the Second written assessment)

 

   

a)     Both means, difference between both means, mean difference for population with 95%CI,  probability value, whether or not 95%CI for mean difference includes zero & what it means in terms of the overall conclusion for the study.

   

b)     Both means, mean difference, interpretation of 95%CI, statistical sig based on p value & 95%CI (whether or not it includes 1), retain/reject null & overall conclusion as to what it means. 

   

c)     Both means, difference between both means, mean difference for population with point esimate,  probability value, whether or not 95%CI for mean difference includes zero, retain/reject null & what it means in terms of the overall conclusion for the study.

   

d)     Both means, mean difference, mean difference for the population with 95%CI,  probability value, whether or not 95%CI for mean difference includes zero, retain/reject null & what it means in terms of the overall conclusion for the study.

Question 12

For all t tests (i.e. One sample, Independent samples & Paired/Dependent samples t test) which one of the following is essential:

 

   

The test variable must be transformed to natural log in all instances.

   

The test variable has to be categorical data (Nominal or Ordinal)

   

The test variable has to be continuous data (Interval or Ratio)

   

Test variable should not have normal or even roughly normal distribution.

   

For the assessment, all assumptions must be met and not violated. 

Question 13

Please identify two incorrect statements from the following. Two answers are required.

 

   

a)     If p value from any assumption test is >.05, it means the study findings could be due to random chance and results are not statistically significant.

   

b)     We use 'average' standard deviation to calculate Cohen's d  instead of 'Pooled' standard deviation as average SD is a good approximation of a pooled value without the complexity. (Heads up for the second written assessment).

   

c)     Before and after types of studies are not true experiments.

   

d)     Time series design, to see if observed trend after an intervention is different from an expected natural trend without the intervention, is considered a type of quasi-experimental design.

   

e)     If 95% confidence interval for mean difference does not include zero we reject null and results have statistical significance and Practical significance.

   

f)      P value of <0.05 for an assumption test means the assumption is violated, whatever that assumption or requirement may be e.g. normality, equal variance etc.

   

g)     Cohen's d is a measure of effect and tells us how far apart the two means are. Cohen's d is equally suitable in situations where we use Independent samples and also where Paired Samples t test is appropriate.

 

NEED IT ASAP. 

WITH SOLUTION IF APPLICABLE!!!!!!!!!!!!

THIS IS Foundations of Biostatistics and Epidemiology

 

Solution by an expert tutor
Blurred Solution
This question has been solved
Subscribe to see this solution