Research methods

Question 1

Studies using experimental methods, but lacking random assignment

Answer 1

Are called quasi-experiments
Are more vulnerable to threats to internal validity than ones using random assignment
Are vulnerable to selection bias

Question 2

Adding repeated pretests to the design of quasi experiments helps reveal

Answer 2

Testing effects

Maturation effects

Question 3

Which of the following design elements helps to reveal selection bias:

Answer 3

Proxy pretest

Single pretest

Question 4

These methods and design elements can be used to prevent or decrease the probability of a selection bias even in the absence of pretest:

Answer 4

Matching
Random assignment
Stratification
Internal control group

Question 5

When including a control group is not possible, it is possible to mimic the function of an independent control group by using

Answer 5

Regression extrapolation contrast
Secondary source contrast
Norm comparison contrast

Question 6

The following design elements can be used to handle maturation effects:

Answer 6

Repeated pre-tests

Question 7

Mark the statements that are true regarding random sampling and random assignment

Answer 7

Random assignment is used to ensure that the groups are randomly similar to each other
Random sampling is used to ensure that sample we take is similar to the population of interest

Question 8

We are conducting a study on clients with GAD. We assign participants randomly to four groups: 1) clients getting 4 CBT sessions over 4 weeks and getting no home assignments. 2) clients getting 8 CBT sessions over 4 weeks and getting no home assignments; 3) clients getting 4 CBT sessions over 4 weeks and getting home assignments; 4) clients getting 8 CBT sessions over 4 weeks and getting home assignment. The following is true for this study:

Answer 8

The study utilize 2x2 factorial design

Data from this study can provide info about the main effect of increasing the number of CBT sessions

Question 9

We are conducting two studies on clients with generalized anxiety disorder. In one of them, we assign participants randomly to two groups: 1) clients getting 4 CBT sessions over 4 weeks and getting no home assignments; 2) clients getting 8 CBT sessions over 4 weeks and getting no home assignments. In the other study, we assign participants randomly to two groups: 1) clients getting 4 CBT sessions over 4 weeks and getting home assignments; 2) clients getting 4 CBT sessions over 4 weeks and getting no home assignments; The following is true for these studies:

Answer 9

Data from these studies can provide information about the main effect of increasing the number of CBT sessions.

Data from these studies cannot provide info about increasing the number of cbt sessions and home assignments.

Question 10

We are conducting a studies on clients with generalized anxiety disorder. If we allocate participants into a treatment and a control group randomly:

Answer 10

We equate groups on the expected value on anxiety scores at pretest

The study can yield unbiased estimates of the average treatment effect

Some threats to internal validity are mitigated

Question 11

We are conducting a study on clients with generalized anxiety disorder. We assign participants randomly to eight groups: 1) clients getting 4 CBT sessions over 4 weeks and getting no home assignments; 2) clients getting 8 CBT sessions over 4 weeks and getting no home assignments; 3) clients getting 4 CBT sessions over 4 weeks and getting home assignments; 4) clients getting 8 CBT sessions over 4 weeks and getting home assignment. 5) clients getting 4 CBT sessions over 8 weeks and getting no home assignments; 6) clients getting 8 CBT sessions over 8 weeks and getting no home assignments; 7) clients getting 4 CBT sessions over 8 weeks and getting home assignments; 8) clients getting 8 CBT sessions over 8 weeks and getting home assignment. The following is true for this study:

Answer 11

Data from this study can provide information about the interaction between increasing the number of CBT sessions and getting home assignment

Question 12

The study design in which participants are randomly assigned to receive either treatment A or B, and following a posttest, they receive the treatment they did not previously get is called:

Answer 12

Crossover- design

Question 13

Which of the following statements are true regarding prospective cohort studies?

Answer 13

They primarily work with data collected over a follow-up period after the study was initiated

They usually take a long time because the researchers have to wait for the event of interest to happen

They include a control group (unexposed group)

Question 14

We are assessing the effects of a new law restricting the prescription and accessibility of opioids in a country. We are interested in the incidence of accidental overdoses causing hospitalization or death. The new law appeared in 2010. In our study (conducted in 2018) we are analyzing past records of the incidence of hospitalization and death due to opioids since 2000 until 2018, which is accessible for each month of these years. The results show that right after the law was appeared in 2010, there was a sudden decrease in the incidence of the events of interest. Furthermore, while before 2010 the incident rates increased steadily over time, after 2010, the incidence rates started to decline. Which of the following statements are true regarding this study and the results?

Answer 14

This study design can be vulnerable to instrumentation threats to internal validity

This study uses an interrupted time series design

The results show a change in slope

Question 15

Which of the following statements are true regarding case-control studies?

Answer 15

They primarily work with data about the past (information about the time before the initiation of the study)

They allow for multiple risk factors to be assessed for one outcome of interest

Question 16

We are conducting a study to understand the risk factors involved in the development of postnatal depression. We identify a group of 100 women who have postnatal depression and select a group of 100 women from the same population who don’t have postnatal depression. We then gather information about partner support during the pregnancy and prior history of depression to see whether these factors are different in the two groups. Which of the following statements are true related to this study.

Answer 16

Case-control study

Question 17

We are conducting a study to understand the effects of childhood cancer on the long term mental health of teenagers. We identify a group of 200 children who have recently underwent treatment for malignant tumor and survived, and we select a group of 200 children who have recently underwent treatment for other types (non-tumor) of acute illnesses requiring more than one month hospital treatment and survived. We then follow these participants until they are diagnosed with depression or other types of mood disorder, or until 10 years (whichever comes first) to see whether there are differences in the two groups in the incidence and timeframe of developing these disorders. Which of the following statements are true related to this study.

Answer 17

Observational study

Prospective cohort study

Question 18

Which of the following statements are true regarding the weights of the studies in meta-analyses

Answer 18

Information about effect size is weighted based on the sample sizes of the studies

Meta-analysis usually examine effect sizes rather than the p-values within studies

Weights are used because the effect size shown on a larger sample size is more likely to be closer to the true mean effect size in the population

Question 19

Which of the following statements are true regarding the difference between fixed and random effect meta-analyses.

Answer 19

Fixed effects analysis is better than random effects meta-analysis if the studies are estimating underlying population parameter

Question 20

We are conducting a meta-analysis on studies which examined the effectiveness of EMDR (Eye Movement Desensitization and Reprocessing) therapy on improving PTSD symptoms. In some of the studies, the origin of PTSD is childhood abuse while in others it is wartime experiences. We suspect that the source of the trauma can influence the effectiveness of EMDR. Which of the following statements are true related to this meta-analysis.

Answer 20

The type of trauma could affect the effect sizes found in different studies

The moderator test would allow us to assess whether study effects vary as a function of trauma type

Taking into account the moderators could partly explain heterogeneity

The type of trauma can be examined as a moderator

Question 21

Publication bias means that:

Answer 21

Some studies are more likely to get published than others

Studies without statistically significant result are less likely to be published

Question 22

Which of the following methods are used in meta-analyses to detect or correct publication bias

Answer 22

Funnel plot

Moderator analysis comparing the effects from published studies with those from unpublished studies

Trim and fill

Question 23

Which of the following methods could be used if some studies that are eligible in the meta-analysis have data from two or more different measures related to each other?

Answer 23

Average the effect sizes within the study and report the mean effect size

Use only one effect size for each study, chose based on an intentional process

Use only one effect size for each study, chose based on a random process

Question 24

In randomized controlled trials (RCTs), the simple randomization

Answer 24

Suppose to ensure internal validity of the research

Is useful to avoid selection bias due to selection effects

Question 25

The treatment as usual (TAU) comparison group

Answer 25

May increase the external validity of the study Can receive various treatment, therefore the meaning of comparison with the specific treatment group is not clear Can hinder the internal validity of the study Is used mainly in effectiveness research Reflects the ordinary care of the people in the treatment setting

Question 26

Differential attrition

Answer 26

Can be a good reason to use intent-to-treat analysis Can decrease both internal and external validity

Question 27

The clinical significance of change can be expressed with

Answer 27

The high proportion of participants whose symptoms improvement met the criteria of clinically significant change (reliable change index > 1.96) The large effect size of change due to the treatment

Question 28

The roles of feasibility studies can be

Answer 28

To evaluation and refinement of data collection procedures and outcome measures To understand if we can recruit appropriate number of participants To estimate cost-effectiveness of treatment

Question 29

Which of the following statements are true regarding retrospective cohort studies?

Answer 29

This is an efficient design for investigating diseases with long latency periods They primarily work with data about the past (info about the time before the initiation of the study)

Question 30

Investigator allegiance bias

Answer 30

Describe the possibilities that enthusiasm of the investigator for a favoured therapy might result in better training and supervision of the therapist implementing the treatment as opposed to a less preferred comparative treatment Can be stronger in studies with low methodological quality compared to high methodological quality Can be stronger if the experimenter was also both the developer of preferred treatment and supervised or trained the therapist Refers to the results being contaminated or distorted by the theoretical treatment preferences of the investigator