chapter 6 Flashcards
mono-operation bias
when only one version of the treatment or program is used. if you use one version of a program in a single place at a single point in time, you may not capture the full breadth of the concept of your program.
solution: try to implement multiple versions of your program
mono-method bias
when you only use one measure of a construct.
solution: implement multiple measures of key constructs and demonstrate that the measures behave as theoretically expected.
interaction of different treatments
when you think what was implemented was the construct you had in mind, but that was not the case
solution: try to isolate effects of your program from effects of other treatments by using a control group.
interaction of testing and treatment
the label program includes testing (testing makes a group more receptive to treatment)
solution: randomly assign your participants in two groups and pretest one of them. if there’s differences in the outcome there’s an effect of pretesting.
restricted generalisability across constructs
when a treatment is only effective on the construct you measured. when you fail to anticipate that the treatment may have negative consequences or side effects on other outcomes
solution: anticipate the unintended and measure a broad range of potential outcomes.
confounding constructs
when the label is not a good description of what you implemented.
solution: if you find a treatment effect at a specific dosage, be sure to conduct subsequent studies that explore the range of effective dosages
social threats
threats related to social and human nature of research
hypothesis guessing
most people engage in hypothesis guessing when participating in a study. they try to determine the real purpose of the study and are likely to base their behaviour on what they guess, and not just your treatment.
solution: control it through research design. eg. have multiple program groups and give each a slightly different explanation. if they perform differently, it may be evidence that it influences results.
evaluation apprehension
influences the way participants behave in a study. eg. when people want to look good or smart, which can not be labeled as a treatment result.
solution: telling participants there are no right or wrong answers.
researcher expectancies
the researcher can bias the results of a study consciously or unconsciously. eg. the researcher can communicate what the desirable outcomes are to the participants.
solution: have multiple researchers who offer different characteristics, if resources allow.
