CASE III

A juvenile correctional institution had implemented a buddy system (pairing each child with a particular staff member) with its younger (6 to 12 year old) juveniles.This system was an attempt to decrease the degree of deviancy exhibited by these children.

To begin, the institution administered the "behavioral deviance" scale (a paper and pencil hostility measure) to their residents, and selected, on the basis of this scale, the 10% of individuals who scored the highest deviance scores and the l0% who scored the lowest deviance scores.

Both groups of participants (high deviance, low deviance) were then exposed to the buddy system, and it was hypothesized that the high deviance group would show more improvement (in behavioral deviance scores) than would the low deviance group.

The hypothesis was supported in that, after exposure to the buddy system, the children who were initially high in deviance had become less deviant (as measured by the hostility measure) while those who were initially low in deviance had become more deviant!The institution concluded that a buddy system decreased deviancy.

a) Research Design:Please describe the research design and explain your choice.

In doing so, label the study as:

• e.g., experimental, correlational, quasiexperimental, ex-post facto, time series etc;

between participants or within participants, or mixed design;

lab or field.

b) Independent variable(s):Please indicate whether there are independent variable(s) and if so, what they are.In describing any independent variable(s),

• please name the specific variable(s)

indicate the levels (i.e., conditions) of each variable and

also describe each variable as manipulated or non-manipulated.

c) Dependent variable(s): Please indicate whether there are dependent variable(s) or any measures. In describing the dependent variable(s)/measures,

• please name the specific variable(s),

label each variable as nominal, ordinal, interval or ratio, and

also explain why you gave the answer(s) you provided in part b.

d) Internal Validity:Please explain any sources of threat to internal validity; also please explain any features in favor of internal validity.(Use 'technical' terms where relevant, e.g., confounding, demand characteristics, regression to the mean, mortality threat, etc., and explain how the terms relate to the situation).

e) External Validity:Please explain any sources of threat to external validity; also please explain any features in favor of external validity.(Use 'technical' terms where relevant.)

f) Statistical Analysis:Please suggest descriptive statistics that are appropriate for the data, and an inferential statistic or statistics that can test the hypothesis.Where appropriate, also indicate whether the analyses are parametric or non-parametric and why those analyses (or that analysis) are (is) are the 'right' ones.