Field Methods pt. 2

within-groups design

There’s only one group of participants and these will be the same participants that will experience all levels of the IV

between-groups design

• Involve two or more groups

• We look at the difference between the two groups

case study

an in-depth investigation + observations of a single individual, family, event, or other entity

● No control/ comparison group

one-group pretest-posttest design

● Cannot determine whether the difference between pre- and posttest is caused by the treatment or a confound

● How sure that the change is solely because of the treatment?

● We do not have a comparison group. We only have a comparison observation.

static-group comparison design

● We are using a pre-established group

● Cannot control for attrition, location, and subject characteristics

● We have two groups:
○ One is subjected to treatment, then observed
○ Other group is not administered the treatment, then observed

● We can now compare two groups;
○ However, this is a static group. They are not randomly assigned.
○ The cause might be because of the groupings, and not the intervention

static-group pretest-posttest design

● Still does not control for subject characteristics and testing threats

● No randomization so we still cannot control for any participant characteristics

repeated measures design

● A within-subjects design

● Vulnerable to testing and order effects

● Participants are measured on DV more than once = We have one group and we give them the treatment on more than one occasion; but also observing them on more than one occasion

■ This gives us a baseline on what their initial behavior is after treatment
■ If there is a significant difference after the second treatment, we could say there is an effect

● There is no comparison group = We are using one group
○ Their exposure to the treatment/ questionnaires more than once might be the effect

concurrent measures design

● Vulnerable to testing effects and demand characteristics

● We have one group and we are administering more than one treatment to them at the same time = concurrent ○ and looking at what the effect is on the participants

● Because there are more than one testing, still vulnerable to testing effects and demand characteristics

maturation

a threat to internal validity where there is change in behavior that emerges more or less spontaneously over time

● People change over time

history

When an external factor systematically affects most members of the treatment group at the same time as the treatment itself, making it unclear whether the change is caused by the treatment received

● Any external factor that affects people in the same direction/ same way

regression

● when a group score is extreme at time 1 and decreases (regresses) at time 2, closer to the typical or average performance

● Occur only when a group is measured twice, and only when the group has an extreme score at pretest

● Happens when we recruit people who have very extreme scores: either extremely high or extremely low

attrition

● Also called mortality

● when a systematic type of participant drops out of the study before it ends (e.g., extreme scorers)

● A participant drops out of the study

● The mean would change at posttest

testing

● A specific kind of order effect

● refers to a change in the participants as a result of taking a test (dependent measure) more than once

● People take the same test more than once

● Change happens because they either got better or got tired

● Practice effect

instrumentation

● When a measuring instrument changes over time

● In observational research, the people who are coding behaviors are the measuring instrument, and over a period of time, they might change their standards for judging behavior by becoming more strict or lenient

● Could also be when a researcher uses different pre and posttest instruments

● We change the way we operationalize or we measure our variables form time 1 to time 2

● Different observers or different questionnaires at the times

weak experimental design

● They do not have built-in controls for threats to internal validity

● Inherently, we cannot ensure that any change in our DV is solely due to IV

true experimental design

● Have a well-designed comparison group

● Randomization is also employed

● Random assignment

randomized posttest-only control group design

Major threats: Attrition (particularly in the treatment group), testing, history, etc.

○ R: randomized members for the treatment & control group

○ X: intervention for treatment group

○ O: a posttest ● Although we do not have a pretest, we do have the control group who, because of the randomization, we are assuming is more or less the same

randomized pretest-posttest control group design

● Major threats: Order (Practice) effects

● Several observations:
○ Treatment & Control @ Pretest ○ Treatment @ Posttest
○ Control @ Pretest-Posttest
○ Treatment & Control @ Posttest

● This is commonly used in experimental research because this is a very strong design
○ But participants are taking the tests more than once, so we could have parallel forms of the same test to avoid threats

randomized solomon four group design

● Ideal but not very practical
○ Controls for most threats to internal validity
○ But needs a larger number of participants

■ At least 30 people per group
■ 120 for the experiment
■ Ensure no attrition happens

● Solomon 4 = there are four groups

● Designed by Solomon Asch

● We have 2 treatment groups and 2 control groups
○ 1st T & C groups - given the pretest
○ 2nd T & C groups - will not be given the pretest

○ Both T groups - will be administered the intervention ○ Both C groups - will not be administered the intervention ○ All groups - will take the posttest