Reference no: EM131039057

Here is some real life data that you are to download from Moodle: file name = takehome2_REG_v3.sav. Your goal is to answer the following research questions from a sample of n = 400 employees at multiple worksites in the Southern California region (spanning biotech, education, and entertainment industries).And as has been the case for all prior homework's and exams, along your statistical conclusion please discuss the implications of your findings and what you will report back to management.

(1) There has been quite a bit of interest of late in the use of technology in the workplace. Given that, I want you to examine how well does the interaction of employees age [Age]and self-confidence [SelfConf] (this measures is based on the mean average of 10 self-report Likert items with mean score ranging from 1 to 6; the higher the score the higher the perceived self-confidence) predict comfort with technology [ComfTech, this is also based on the mean average of 10 Likert items with mean score ranging from 1 to 6---the higher the scale score the more one feels comfortable with technology, such as internet, skyping, smartphones, etc.]. Conduct a test of the interaction in SPSS and provide the appropriate graphics (examine assumptions such as normality of residuals and homoscedasticity) and furnish a detailed interpretation. What are your conclusions (don't forget to keep in mind not only significance but also effect sizes)? And as excerpted from the homework:

Note: If the continuous x continuous interaction is significant, even though you aren't required to graph such, one means by which you can interpret the interaction (as shown in class) is to segment the moderator (in this case it will be age) by three categories: cases 1 SD or below the mean, cases between +1SD below or cases that are above -1 SD below the mean and then cases that are above +1 SD. So for example, let's say age has mean of 50 and SD of 10, you could create the following categories:

(1) age< 40 (one SD below the mean)

(2) agebetween 40 and 60 (between -1 SD and +1 SD)

(3) age> 60 (one SD above the mean)

And then you can create the scatterplot with the ‘marker' being the age_Category variable (three categories), and y-axis = ComfTechand x-axis = SelfConfAgain, this is not required, but would be helpful to aid interpretation (even if not significant) and as we went over in the homework, using the split file function and running the simple linear regression for each age category (self-confidence - > comfort with technology) will also aid in interpretation.

(2) Research has shown that leaders may be inclined to operate from a desired power base, three of those being as follows (from: https://quickbase.intuit.com/blog/2011/08/26/the-5-types-of-power-in-leadership):

Coercive

Coercive power is conveyed through fear of losing one's job, being demoted, receiving a poor performance review, having prime projects taken away, etc. This power is gotten through threatening others. For example, the VP of Sales who threatens sales folks to meet their goals or get replaced.

Reward

Reward power is conveyed through rewarding individuals for compliance with one's wishes. This may be done through giving bonuses, raises, a promotion, extra time off from work, etc. For example, the supervisor who provides employees comp time when they meet an objective she sets for a project.

Legitimate

Legitimate power comes from having a position of power in an organization, such as being the boss or a key member of a leadership team. This power comes when employees in the organization recognize the authority of the individual. For example, the CEO who determines the overall direction of the company and the resource needs of the company.

I am interested in seeing how regression parallels ANOVA as discussed in class. Thus, please compare the groups in the variable capturing 'type of power' labeled Power_Type via dummy coding, and use the third level ('reward') for power type as the reference category for the dummy coded vectors. The criterion variable of interest is job satisfaction (JobSat_SC: higher the score the higher the job satisfaction). Run both regression (with the dummy coded vectors) and one-way ANOVA and compare the results. Please interpret the regression output and compare to ANOVA, and offer your conclusion.

(3) Next, you are interested if the relationship of positive affect1 and jobsatisfaction depend upon the preferred power type. Thus, you are going to test the following model for this assignment:

How well does the interaction of type of power (three categories) and positive affect (higher the value the higher the positive affect[pos_affect].and don't forget to center!!) predict jobsatisfaction (the higher the value the higher the job satisfaction for job_Sc). Conduct a test of the interaction in SPSS and use the third level (reward) of power type as the reference category for the dummy coded vectors. What are your conclusions? And then I want you to run the simple linear regressions for each level of power type (positive affect predicting job Satisfaction) and in your opinion do the pattern of coefficients seem to correspond with the conclusions you have drawn from themoderated multiple regression? Also whether significant or not display the appropriate graphic of fit lines as discussed in class so as to interpret the interaction term (and best to use uncentered predictor for graph so as to aid interpretation of the raw score).

Positive Affectivity is a characteristic that describes how animals and humans experience positive emotions and interact with others and with their surroundings.[1] Those with high positive affectivity are typically enthusiastic, energetic, confident, active, and alert. Those having low levels of positive affectivity can be characterized by sadness, lethargy, distress, and un-pleasurable engagement (see negative affectivity).

(4) And with much press (and peer reviewed literature) being paid to the convergence of health, wellness, and psychological factors (and the attendant impact in the workplace) your final task is to run multiple regression to see how wella self-rating of General Health (Gen_Hlth: average of Likert items with values ranging from 1-6, a higher score indicating higher rating of one's generalhealth), Social Support (SocSup-this self-report scale, also averaged items from 1-6 measures to what extent the respondent is satisfied with their level of social support, such as from colleagues, family, friends, etc.with a higher score indicative of more satisfaction with their social support system) and average hours sleep a night (Sleep) predictPositive Affect (Pos_Affect) which has already been discussed. That being said, run a simultaneous multiple regression, summarize the results and specifically discuss the assumptions from chapter 10, such as outliers, influence, and collinearity. And of course, put it all together with some of the graphics you learned in chapter 4 (linearity, normality of residuals, residual plot).

Attachment:- takehome2_Reg_V3.rar