K205 Module 8: Quantitative Research Design

FHV: Define what a clinical trial is. Comment on control group

Any trial where we prospectively assign human participants to a "health related intervention” and measure health related outcomes

• No control group since not ethical

FHV: Distinguish between a randomized control trial and clinical trial 

RCT:

• Randomly assigns participants to control or intervention group and does pre + post testing

• Random assignment of who got put in which group (equal chance of being put in each group)

Clinical trial:

• any trial where we prospectively assign human participants to a health related intervention and measure health related outcomes

FHV: Understand why researchers use prospective registration

Includes:

• Inclusion and exclusion criteria

• Sample size

• Primary outcomes are important so that one can’t rig the study outcomes if it didn’t work out as expected

FHV: Understand the purpose of a flow diagram

• Shows the flow of research design from beginning to intervention

• Show how many people were interested in trial, how many met inclusion criteria, how many lost, how many finished

• Important because:

  • future researcher can change inclusion criteria based on how many people signed up compared to how many wanted to sign up 

  • find out if an intervention, even if effective, was so hard that it caused lots of attrition or injury

Differentiate between, identify and understand the purpose of non-experimental and experimental research designs

1. Non-experimental:

   • Theoretically presumed cause and effect

   • No intervention

   • No control group or random assignment

   • Observational design or correlational design

     • Observation: How? what?

     • Correlational: the strength of the association between two variables

       • Positive or negative direction of relationship

2. Experimental:

   • Cause and effect = determinism

     • manipulate IV(s) to determine if it has an effect on DV(s)

     • Intervention

     • Random assignment and control group

Identify cross-sectional and longitudinal research design

Cross-sectional: Only take one measurement, at one time point. It can be for multiple variables

• Cannot tell us about change over time

Longitudinal: Repeated measurement of variables over time (weeks, months, years etc) 

Identify between-groups and repeated measures designs

Between-groups design:

1. Separate group for each condition

2. Participants only provide data once. Only one observation 

Repeated measures: 

1. Observations taken from same participants more than once 

2. IV/Predictor = ‘Time of observation’

3. One group 

4. More than one observation 

5. One treatment

Differentiate between and identify pre-experimental, true experimental and quasi-experimental research design

Pre-experimental: 

1. no random assignment 

2. They suck because they lack internal validity

3. Changes in DV cannot be attributed to manipulation of IV

4. Participants receive treatment of interest 

True Experimental: 

1. Random Assignment!

2. Helps with: history, maturation, testing

Quasi-experimental: 

1. Interested in maximizing external validity; want to approximate real world settings 

2. No random assignment possible 

3. Participants ‘self select’ to one of the groups OR the administrator (coach for ex) decides who receives the treatment 

Identify and describe the strengths and limitations of 3 pre-experimental designs 

1. One shot design: One treatment, one observation afterwards, one group

Group A: T → O₁

Strengths: 

• Inexpensive 

Limitations: 

• No internal validity because we don’t know what happened to participant before the treatment that could affect the results

2. One Group Pretest-Posttest study: 

   • Happens with lots of S&C studies 

   • Goal: try to determine the magnitude of the treatment effect 

   • Repeated measures design 

Group A: O₁ → T → O₂

Strengths:

• Pre-test helps us compare results pre and post

Limitations:

• Cannot attribute any change in performance to the intervention

3. Posttest Only with Non-Equivalent Groups

N Group A: T → O₁

N Group B: O₂

• Between-groups design

Strengths: 

• Has a control group so can compare post test results 

Limitations:

• Groups are NOT randomly formed (intact groups)

• Pre-existing differences between the two groups, especially the treatment group, causing it to have higher or lower results

Identify and describe the strengths and limitations of 3 true experimental research designs

True experimental designs have random assignment!

Posttest only Control Group Design

Group A: T, O

Group B: __, O

Strengths:

• Random assignment to treatment and control group helps account for selection bias 

Limitations: 

• No pretest

Posttest only Control Group Design with 2 treatments: 

Group A: T1, O

Group B: T2, O

Group C: __, (

Strengths: 

• Determine if one treatment is more effective than another and if it is better than nothing

Limitation: 

• No pretest 

Pretest Posttest Control Group Design:

Group A: O, T, O

Group B: O, __, O

Strengths: 

• Added pretest to compare measurements of variables. Increases internal validity

• Determine how much change is observed in  Group A vs Group B

Limitations: 

• Threat to external validity → reactive/interactive effects of testing 

• Hard to generalize finding

Solomon group 4 design: 

• Group 1: T, O

• Group 2: _, O

• Group 3: O, T, O

• Group 4: O, _, O

Strengths: 

• Combines post test only and pretest post test designs together

• Helps to see if there is a difference post test for those who did a pretest or not. 

• Helps to account for the reactive/interactive effects of testing (where treatment may not be as effective without the pretest)

Limitations: 

• Very expensive to repeat treatment.

Identify and describe the strengths and limitations of 3 quasi-experimental research designs

…