Study Designs

Name this study design

100

Involves collecting data to determine to determine whether and to what extent a relationship exists between two variables (associational research). Variables are not manipulated and random assignment is not possible. Purpose is to discover or clarify relationships between variables and assessing this best by utilization a correlation coefficient (+1 to -1). 

Correlational research 

200

In this type of study, when well conducted, every learner has an equal and independent chance to be assigned to intervention or control group which protects against the possibility that factors other than the intervention itself will influence or confound the results which is a main determinant of the validity of the results. 

RCT

• Additionally assignments to groups should be concealed from those who administer the assignments at the site. 
• Solomon four group design: two groups which receive intervention and two which do not, and combines varying observations with the presence or absence of treatment, moderating the Hawthorne effect when improvement occurs as an artifact of participation only 

300

For this type of study, it is important to: form a review group, write a review protocol, manage the citations and abstracts and primary studies, search the literature and data management, rate the quality of primary studies and measure effectiveness of interventions, identify relevant search terms and thoroughly search relevant databases

Systematic Review 

- preferably 2 or more review group members should read all potentially relevant abstracts in order to ascertain whether they meet inclusion criteria for the study

- One useful metric in these medical education studies is the number needed to teach which refers to the number of learners who need to be taught by the intervention of interest in order for one learner to benefit. 

Example of study design

100

a survey of psychiatry chief residents to determine self-reported adherence of training programs to the US accreditation Council for GME standards for emergency psychiatry training. 

Survey or descriptive research 

200

analysis that included 59 studies that enrolled a total of 8,011 learners and demonstrated that self-directed learning is associated with moderate improvement in the knowledge domain compared with traditional teaching methods 

Systematic review

300

Example: Students entering medical school in 2011 were allowed to choose between one of two intervention groups: discussion or self-help. These two non-randomized groups were compared to a control group consisting of students who entered medical school in 2012. Researches demonstrated that the intervention had a positive effect on perceived med student stress due to participation in self development groups. 

Quasi-experimental 

Name this study design

100

This type of study is used to determine whether there is a cause and effect relationship between two variables. It usually starts with an effect and seeks possible causes. 

Causal Comparative research 

- The advantage of causal comparative research design is that it allows the study of cause and effect relationships under conditions where experimental manipulation is difficult or impossible. 

200

The purpose of this research is to describe existing characteristics of a sample population but the researcher cannot manipulate variables.

Survey or descriptive research 

• This requires use of survey instruments to collect data and requires a sample of individuals which is representative of the population being tested. 
• Quality of this research is dependent on quality of survey instrument, quality or representativeness of sample and response rate (lower response rates contribute to potential for bias). 
• Survey instruments should measure what they purport to measure (validity) and they should measure the variable of interest consistently (reliability) 
• Information can be collected from a group of individuals at one time or over a period of time (cross sectional vs longitudinal survey)

300

In this type of study, there is a control group but it lacks true random assignment. This study is useful to determine a cause-and-effect relationship when randomization or matching is not possible

Quasi-experimental 

• Advantage is that researcher is able to manipulate the independent variable however the disadvantage is potential for significant differences between groups. The lack of random assignment increases the risk that factors other than the experimental intervention will affect the outcome of the study.

Example of study design

100

Grady-Weliky et al described the effects on psychiatryc residents’ self-assessment of teaching knowledge and skills after a brief residents-as teachers workshop. The researchers demonstrated that a brief educational workshop enhanced psychiatric residents’ self-assessment of teaching knowledge and skills. 

Case series 

200

One study identified 40 ‘problem residents’ over a 20 year time period and 40 matched control residents. The researchers retrospectively examined whether data available at the time of residency application could be used to predict future problems of performance, both during and after residency, and found that presence of any negative comment in MSPE was correlated with future problems. 

Causal Comparative research

300

Determine whether or not there was a relationship between the total clinical grade of medical students in a psychiatry clerkship and NBME shelf examination scores. The grade received by students on rotation did not predict shelf score. 

Correlational research