Unit 10 Flashcards
How are quasiexperimental different from two experiments?
Quasi-experiment resembles true experiments but lacks at least one of its defining characteristics. Where it is possible to assign subjects to conditions in a true experiment, often in a quasi-experiment it is necessary to select subjects for the different conditions from previously existing groups
The independent variable in a quasi-experiment is often called a subject variable if it is a characteristic of the subjects on which they have been selected, such as sex of participants
Quasi-experiments are often conducted after the groups have been formed, for instance with the subject variable of sex as the independent variable, the experiment takes place long after the subjects become males or females
Incher experiments we manipulate variables, whereas in quasi-experiment we observe categories of subjects. When we take to pre-existing groups and consider a difference between them to be the independent variable, we are not manipulating a variable but simply labeling groups according to what we think is the important difference between them
The quasi-experiment does not have as much control as a true experiment because it leaves open the possibility that other differences exist between experimental and control conditions
Quasi experiments are sometimes called this kind of experiment because it is conducted after the groups have been formed
Ex post facto, or after the fact
What is the major weakness of quasi-experiment?
Often times in quasi-experiments that who of the experiment cannot be controlled. In some instances the what, when, where, and how cannot be completely controlled either. If enough compromise of experimental control takes place, this lowers the internal validity of the experiment.
What specific situations are quasi-experiment required?
Quasi experimental research allows the examination of natural experience, which are formed when real world events divide people into two groups for comparison. In general, that you experiment is preferable to the quasi-experiment, but many situations existing which randomly assigning subject to conditions is not possible. Then a quasi-experiment perform simply because doing it is better than not doing experiment.
Research design having both an experimental and control groups wherein subjects are not randomly assigned to groups
Non-equivalent control-group design
The problem with this design is in determining how to compare results between experimental and control groups when the two groups were not equivalent to begin with
Under what circumstances are the results of a quasi-experiment using a non-equivalent control group interpretable?
Depends on whether the pattern of results obtained can be accounted for by possible differences in groups or by something else in the experiment.
Results are more interpretable when the two groups to show the same performance on the pretest, the experimental group improved on the post test, but the control group did not change. Their performances can be compared and the results interpreted because their behavior as measured by the dependent variable was the same at the beginning
Another pattern of results that is usually interpretable is when the experimental group was lower than the control group on the pretest but higher than the control group on the posttest
Pretests and post tests that show that the control group was superior to the experimental group but did not improve are not interpretable because it may indicate a ceiling effect, the control group could not perform any better. Also uninterpretable, are the results that show improvement for both groups on the post test
Sometimes no control group can be obtained that can be considered comparable enough to be useful, in this case a design that allows the same group to be compared overtime can be used. What are two such designs?
The interrupted time-series design and the repeated-treatments design
Research design that allows the same group to be compared overtime by considering the trend of the data before and after experimental manipulation
Interrupted time-series design
Design in which a treatment is withdrawn and then presented a second time
Repeated-treatment design
In developmental research, a study that tests different age groups at the same time
Cross-sectional study
The advantage of the cross-sectional approach is that all the age groups can be tested at the same general time. The disadvantage is that the people at different ages were all born at different times, so age is confounded with date of birth
In research, a group that has something in common, such as age
Cohort
In developmental research, a study that tests individuals in a single cohort over the course of time
Longitudinal study
Helps to avoid cohort affects. The advantage of this approach is that all the people would have the same birthdates, so there would be no cohort effects. There are two main problems, practically, the research has to wait years to complete the study as the cohort ages. Theoretically, it confounds age with time of testing. All the participants will be eating together, but at the same time the world would be changing
Design used to help separate developmental, cohort, and secular effects
Cross-sequential design
Tests individuals from two or more cohorts at two or more times
A set of techniques for determining the effectiveness of a social service program
Program evaluation
Quasi experimental designs are often used in programming valuations, because clients of the program are more likely to be assigned, rather than to be randomly allocated two groups for comparison.
And evaluation of the quality of the project, often after it is completed
Summative evaluation
