Tag: Homework Solution

Assignment Prompt & Instructions

Your assignment this week will cover the course material from Week 3 and Week 4.  In this assignment, you will: (1) download the Human Resources (HR) dataset from our course page in Canvas and import it into your chosen software [SPSS, R, or SAS], (2) produce various descriptive statistics for several HR dataset variables, (3) conduct binomial tests on the proportion of employees who left the company, and (4) demonstrate your understanding of the normal distribution and z scores.

Please respond to the following items below in a single Word document (written in Times New Roman 12-point font, double spaced) and submit the single document through the Canvas assignments page.  The assignment submission must conform to the APA-style report format and your responses must be supported with in-text references to the course material (readings, lectures).  Please refer to the assignment rubric for additional details about how your assignment will be graded.

– Produce a bar chart demonstrating the counts for those who filed a complaint compared to those who did not file a complaint (i.e., two bars).
– Calculate and compare the mean, median, and mode of the performance score for those who have filed a complaint (this will involve selecting only those who have a value of 1 for “filed_complaint”; in SPSS this could involve use of the select cases command). What do each of these measures of central tendency reveal to you about the “average” level of performance among those who have filed a complaint?
– Calculate and compare the mean, median, mode, standard deviation, kurtosis, and skewness of the performance scores for those who have filed a complaint vs those who have not filed a complaint (this will involve analyzing those who have filed a complaint separately from those who did not file a complaint; in SPSS this could involve use of the split file command; in R and SAS, this could involve other comparable commands).  What differences do you observe, if any, for each of those results when eyeballing them (as opposed to testing for differences via a statistical test, which we will discuss later)?
– Produce a frequency distribution (i.e., histogram) of performance scores for each group, i.e., those who have filed a complaint vs those who did not.  You may copy and paste the histograms from your output into the Word document (the chart does not have to be formatted in Times New Roman 12-point font).
– Next, consider the variable “left” with left = 0 representing those employees still with the company and left = 1 those employees who left the company (for various reasons).  Determine the observed proportion of those who have left from the HR dataset.
– Conduct an initial binomial test with the test value (or test proportion) set to exactly the observed proportion of those who have left in your sample.  Record the output of the binomial test in your Word document, specifically noting the reported statistical significance of the test (in SPSS this is the "Exact sig" value).
– Then, conduct at least four additional binomial tests with new test values (i.e., test proportion) that are slightly above (at least two tests) and slightly below (at least two tests) your observed proportion of those who have left.  Record the outputs of these additional tests in your Word document.  Based on the readings this week and what you observed from conducting the initial and additional binomial tests, explain and interpret the results you observed (e.g., you might discuss "How sensitive the results were to your slight changes in the test value?" and "Why you think this might have occurred?").
– Next, consider the variable "last evaluation" in the dataset.  The "last evaluation" variable indicates how valued the employee was at the company with raw scores ranging from 0.0 (not valued at all) to 1.0 (very valued).  Produce a set of z scores for the “last evaluation” variable and then produce descriptive statistics to report the mean and standard deviation of the standardized z scores.  Copy and paste the descriptive statistics table into your Word assignment.
– Then, consider the first two employees in the HR dataset (i.e., the first two entries in your dataset).  When considering only their original “last evaluation” score ranging from 0.0 to 1.0 (i.e., not using the associated z score), what do you conclude about each employee's value at the company?
– Next, use the z-to-p calculator at the end of the last Vassar Stats reading in Week 3 to report the following about the “last evaluation” z scores for the first two employees of the HR dataset: (1) the proportion of the distribution that falls above (or below if the z score is negative) that value (this is the one-tailed for -z or +z), and (2) the percentage of the distribution that falls above and below the |z| (this is the two-tailed value).
– When examining the first two employees’ z scores for “last evaluation”, what can you conclude about where each employee's score falls relative to the rest of the sample?  Specifically, what proportion of the sample is less valued than each of these employees and what proportion of the sample is more valued than each of these employees?