To determine if halfway houses improve the transition from prison, 50 inmates wh
ID: 3502398 • Letter: T
Question
To determine if halfway houses improve the transition from prison, 50 inmates who will be released in a month are randomly assigned to one of two groups. One group is sent to a halfway house to participate in a program called "ReEntry." The other group stays in prison until the time of release. Two months after release, data measuring "success in job place- ment" are gathered on the subjects in both groups a Use the R, X, O notation to sketch the research design. b An analyst says that the design does not control for statistical regression since the 25 subjects are inmates and thus represent an "extreme in their social behavior." Do you agree that the design has problems with statistical regression? Explain c Define the terms internal validity and external validity in the context of this study.Explanation / Answer
Note: This response is in UK English, please paste the response to MS Word and you should be able to spot discrepancies easily. You may elaborate the answer based on personal views or your classwork if necessary.
(Answer) (1) R = Random Assignment to a group – The details of the experiment do not seem to mention any particular basis on which the inmates are divided other than the fact that they have the similar release date.
X = Treatment – The research design is purported to measure the transition into normal life after time spent in prison. The control group and the experimental group are varied on the basis of their stay in the halfway houses.
O = Observation – One of the greatest aspects of normal life would include being accepted by society and given a job that would help them regain control of their personal and professional lives. The stay of the inmates in halfway houses and in prison will be compared in terms of the outcome they produce. The benchmark of this outcome will be measured based on whether or not they are able to acquire decent jobs or not.
(2) Statistical regression occurs when researchers tend to include an “extreme” subject in order to manipulate the test results. For instance, there is a research to test the effects of almonds on the IQ of children. If the researcher has a bias and wants to prove the hypothesis to be true, they might take extremely clever subjects to begin with.
If this were a normal experiment, it would be termed as statistical regression to take only prison inmates. However, the context of the study should essentially be considered here. The purpose is to measure a particular “solution” on a particular type of “subject.” Therefore, using only prisoners is an essential criterion. If all the prisoners in the experimental group were petty thieves and the prisoners in the random group were murders, then it would be statistical regression. Since they are randomly distributed into groups, it is not a case of statistical regression.
(3) External Validity – The external validity is tested through two different circumstances. This means that there are two groups of people who are tested. The people from the halfway houses and the ones from prison are the two types of testers. This would enable the experiment to be tested in two different circumstances and not just one single type of test. This increases its external validity.
Internal validity – Maturation is one of the aspects of internal validity that is tested here. This means that there is a different level of maturity in the halfway house prisoners and the cell inmates. Furthermore, the history of each of the prisoner might be different and might be extraneous variables that are not considered. A random assignment might help supplicate these validity threats. However, it would be better if, for instance, a thief from the house and the cell were both tested in a group. In other words, the test should be conducted with a crime or personality counterpart in each group to make sure that there are no there are no other variables at play.
Related Questions
drjack9650@gmail.com
Navigate
Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.