Correlation & reliability using SPSS

Worth (15 Marks)

1. Reference reliability: Data file = 2011 R eference reliability.sav
   This file contains reference ratings for job applicants. Each applicant has two references. For each applicant each reference was asked "Would you hire this person again if he/she applied for a position with your organization?" The references indicated their willingness to rehire the applicant using a 1 (definitely would not rehire the person) to 10 (definitely would rehire the person). The question is whether the references’ responses are consistent?
2. test versions A to E reliability; Data file = 2011 v ersions A to E.sav
   This file contains the results of five different versions of a mechanical abilities test that were completed by people applying for company sponsored trainee position with technical colleges across Canada. Scores have a theoretical range of 0 to 100 where higher scores indicate greater mechanical ability. The question is whether the results people obtain on the different versions of the test are consistent over the five versions of the test?
3. Integrity Testing Reliabilty Part A, cronbach’s alpha: Data file = 2011 HONESTY SCALE.SAV
   This file contains the results of a study to develop a test of honesty and integrity. The data consists of 79 people’s responses to the following set of seven questions.
   1 2 3 4 5
   Strongly Neutral Strongly
   Disagree Agree
   1. It is okay to steal if the item is worth less than a couple of dollars.
   2. Everyone is dishonest to some degree.
   3. If I observed someone stealing at work I would not report them if the items stolen were not too valuable.
   4. Some of the people I know may be dishonest, but friendship is more important to me.
   5. If someone is in need the shouldn’t be held as accountable as someone who steals, but is not in need.
   6. I agree with the saying "Finders keepers, losers-weepers."
   7. If someone offered to give me a really good deal on an expensive item that was stolen I would take the deal.
      The question is whether the set of 7 questions, as a group, provide an acceptable measure of integrity as indicated by Cronbach’s alpha (i.e., ). In other words, do the set of 7 questions represent a group of homogenous measures that can be averaged to create a single overall score representing a job candidate’s integrity and honesty?
      The equation for Cronbach’s alpha is provided below where n = the number of questions in the questionnaire and \[\bar{r}\] = the mean correlation determined from the correlations among the seven questions (you will need a calculator to do this).
      \[\alpha =\frac{n\left( {\bar{r}} \right)}{1+\bar{r}(n-1)}\]
4. Integrity Testing Reliabilty Part B: Data file = 2011 i ntegrity testing.sav

Based on the assessment of Cronbach’s alpha a final set of questions were retained to construct an honest test. The honest test was given to job applicants and an honesty score was determined as the sum of the responses to the questions. The file contains the integrity test scores of job applicants obtained at the time they applied for a position and their scores on the same test completed 6 to 8 months after they were hired. The question is what is the nature of the relationship between the pre and post hiring scores?

Price: $14.44

Solution: The downloadable solution consists of 8 pages, 644 words and 11 charts.
Deliverable: Word Document