SOCW 6311 Selection of a Statistical Analysis Approach

WHAT YOU WILL READ AND/OR WATCH THIS WEEK:

AND Choose 1 or 2:

  • 1. Study the section on the chi-square test of independence which we use in this discussion (Stocks, 2010 in B. Thyer (Ed.), The handbook of social work research methods (2nd ed.), pp. 115-116) The chapter is accessible as a PDF in the list of learning resources.   OR
  • 2. Watch this 17-minute video that explains chi-square tests in the readings:  StatsCast-What is a Chi Square?https://www.youtube.com/watch?v=mFIvsqVir48&t=63s

_________________________________________________________________________________

WHAT YOU WILL LEARN THIS WEEK:

1.  You will become acquainted with a set of statistical tests used in social work research (Stocks, 2010 in Thyer)

2. You will review confounding variables and threats to validity that can weaken or invalidate a study.

3. You will explore one statistical test in detail, the chi-square test of independence, and learn how to interpret the results.Information is presented in the optional reading below the discussion questions.

___________________________________________________________________________________________________________

WHAT YOU WILL DO THIS WEEK:

By Day 3 (Wednesday, 11:59 pm Mountain time, 1:59 am Thursday Eastern time) explain whether you agree or disagree with the researchers’ conclusion from their study on the effect of an intervention on employment for recently paroled clients. Include a comment on potential threats to validity that could impact the study. Please title your submission “Original post”.

_____________________________________________________

Note: In this case study, the social workers evaluated the employment status of recent parolees who participated in a new type of employment program for parolees and recent parolees who were placed on a waiting list.

To make reasonable conclusions, we have to make some important assumptions that were not included in the case study. 

We have to assume that:

1. program participants enrolled on a first-come-first-served basis; they were not specifically selected for the program by any sampling method, including random selection.

2. none of the program participants or waiting list members were employed at the start of the program. In other words, all 60 individuals – program participants and those waiting to enter the program – were unemployed when the program period started.

3. most of the client characteristics were the same for both groups. For example, type of crime, intellectual capacity, physical ability, motivation, mental or chemical health conditions, and other important characteristics were found in both groups and therefore had no effect on the outcome.

4. information on the employment status of the 60 members of the study (treatment and waiting list group) was gathered some time after the program ended to make it possible to compare sustainability of employment. In other words, parolees who took a position but quit shortly afterward would not be categorized as employed.  The employment outcomes that were counted included full-time employment, part-time employment, and no employment (unemployed).

At the end of the program, the employment outcomes for the two groups (participants vs. the waiting group) were compared. Data for each group were entered into a statistical program (SPSS) that produced the following output charts and tables presented here.

An explanation of chi-square analysis and the meaning of the items in the charts are in the optional reading section beneath this set of instructions for the discussion post.

_________________________________________________________

By Day 3, (Wednesday night, 11:59 PM) submit an original post in the format provided here.

Please label your entry “Original Post”. If you decide to revise your post, please label it “Revised post”.

1. Enter your interpretation of the conclusion made by the researchers that “the vocational rehabilitation intervention program may be effective at promoting full-time employment”, by filling in the blanks where indicated in the sentences below.  Explain whether you agree or disagree with the researcher’s conclusion that the program may be effective (notice that she does not say “is effective”) based on the statistical results, and provide your rationale, citing the chi-square statistic from the table above. (2 sentences)

Choose one of these sentences (a or b) for your response:

a1. The researcher has concluded that the vocational rehabilitation intervention program may be effective at promoting full-time employment. I [select one: agree/disagree] with the researcher’s conclusion because the chi-square statistic is ______”. [You may add more information here.]

OR

b1.The researcher has concluded that the vocational rehabilitation intervention program may be effective at promoting full-time employment. I agree with the researcher’s conclusion because the chi-square statistic is ______”, but I have questions regarding these aspects of the study: ___, ___, and ____ because ______________ [You may add more information here]. These problems would have to be addressed before I could give the program my full support. “

[Note: Whatever choice of (a1) or (b1) you make can be valid, depending on your reasons. Provide a plausible and defensible response. Qualified scientists disagree with one another for solid reasons; so can you.]

2. Name and describe TWO factors that could limit the internal validity of this study, using the lists adapted from the Boston University Public Health Program Evaluation webpage describing threats to validity or other sources.

At least one threat must be one that has not been described previously in this discussion forum; the other may not have been discussed more than once previously. Do not use the same factors described by a fellow scholar who posted earlier, if possible.

Ensure that the factors you select are relevant to this study. Maturation, for example, is most common in developing children, but is not expected in adults, so it is not relevant in this study. Mortality is not relevant in this case, either.

You will note that there is one person’s data that was lost, but when a study includes valid data from 59 participants, the loss of data from one participant is not significant and is not a threat to validity in this case.

Use this sentence to respond to Question 2:

2. Two threats to the internal validity of this study are 1)__________, defined as ________ and 2) ___________, defined as________, as applied to this study.

3. Explain why the factors you selected in question 2 limit the ability to draw conclusions regarding cause-and-effect relationships between the independent variable and the outcomes. Describe how you might mitigate the problem caused by the threat to internal validity. (3-4 sentences)

Use these sentences to respond to Question 3:

3. [The item you chose for internal validity #1] limits the ability to draw conclusions regarding the effect of the intervention because ___________. One approach to revise the study to eliminate or mitigate the problem would be to ___________ I would suggest…etc.”

[The item you chose for internal validity #2] limits the ability to draw conclusions regarding the effect of the intervention because ___________. One approach to revise the study to eliminate or mitigate the problem would be to ___________ I would suggest…etc.”