Quasi-Experimental Research Strategy

Chapter 10

Nonequivalent group design

• the researcher does not control the assignment of participant groups

  • the groups already exist (ex; age, gender…)

  • problem with internal validity (lack of control)

Assignment Bias

• occurs when the comparison group participants differ in some systematic way(s) from the treatment group participants

  • ex; differ in intelligence, race, age, gender, family background…

Quasi-Experimental Research Designs

• designs that approximate the features of a true experiment

  • “quasi” meaning similar to (or resemblingg)

3 Common Examples of Nonequivalent Group Designs

1. differential research designs

2. nonequivalent control group design

3. pre-post designs

Differential Research Design

• a study that simply compares pre-existing groups (simply determine is there a difference)

  • e.g., males vs females

Nonequivalent Control Group Design

1. posttest-only nonequivalent control group design

2. pretest-posttest nonequivalent control group design

Posttest-only nonequivalent control group design

• uses pre-existing groups, one of which serves as the treatment group and the other as the control group

  • one group of smokers volunteers to get treatment

  • one group of smokers does not volunteer to get treatment so they become the control group '

  • both groups complete measure of cigarettes smokes after a week (with or without treatment)

Pretest-posttest nonequivalent control group design

• compares 2 pre-existing groups (i.e., assignment is not random)

  • one group is measured twice (before and after the treatment is given)

  • the other group is measured at the same times but does not receive any treatment

Advantages of pretest-posttest compared to posttest only

• the pretest scores allow us to see how similar the 2 groups are before treatment is administered

  • if the 2 groups are similar, it reduces (but does not eliminate) the threat of assignment bias

• can look at changes in scores from the pretest to the posttest

Pre-post designs

• time-series design

• interrupted time-series design

• one-group pretest-posttest design

Time-Series Design

• a study in which a series of measurements are taken both before and after a treatment manipulated by the researcher has occurred

  • ex; number of bullying incidents at a school for several weeks before and several weeks after an anti-bullying program

Interrupted Time-Series Design

• a study in which a series of measures are taken both before and after the occurrence of some naturally occurring event (ex; mega construction project, earthquake…)

  • ex; track PEI tourism for years before and after the construction of the bridge to ensure it wasn’t the celebration about the bridge or something else that caused the increased tourism

• helps us to rule out temporary factors like media hype

• does not help us rule out history effects

One-Group Pretest-Posttest Design

• a study in which only one measure is taken in both before and after a treatment or natural event has occurred

  • ex; look at PEI tourism the year before and after the bridge was built

• does not account for time-related variables

  • ex; maybe there was a big tourist event the year after the bridge was built so the improved tourism was related to the tourist event not the bridge itself

Developmental Research Designs

• cross-sectional design

• longitudinal design

Cross-Sectional Study

• a study of groups of individuals of different ages at the same time

Advantages of Cross-Sectional Studies

• less expensive

• get your results right away (i.e., within a relatively short period of time)

Disadvantages of Cross-Sectional Studies

• cohort effects: differences among groups of individuals of different ages that result from generational effects (e.g., different economic, social, political, educational, and family circumstances)

Cohort

• a group of people who were all born at about the same time and grew up under similar circumstances

Longitudinal Studies

• the study of the same group of individuals at different points in time as they grow older

Advantages of Longitudinal Studies

• only way to conclusively study changes that occur as people grow older

• best way to study how scores on a variable at one age are related to scores on another variable at a later age

Disadvantages of Longitudinal Studies

• expensive

• time consuming

• participant attrition: participants dropping out of a study (e.g., because they move away, lose interest, die, etc.)