Data and Analytics Exercise: Evaluating Employees Satisfaction With Benefits Offering the right mix of benefits can be quite challenging. Employers want to manage costs while also providing benefits that attract, motivate, and retain potential and current employees, and employees want benefits that meet their needs. Even the best assortment of benefits offerings can face problems if program details are not properly communicated and explained to employees. Employee surveys can be a powerful data collection tool when used to understand employees attitudes toward benefits and how to improve benefits offerings. Imagine an employee survey with items (e.g., questions) pertaining to the following attitudinal and behavioral concepts: overall benefits satisfaction, turnover intentions, and attendance at a benefits information session. Overall benefits satisfaction and turnover intentions are assessed with five-item measures, where employees rated each item using a 1 = strongly disagree and 5 = strongly agree response scale. An example item for overall benefits satisfaction is: I am satisfied with the companys current medical plan offerings. A sample item for turnover intentions is: I am considering leaving the organization in the next 6 months. A single item is used to assess attendance at a benefits information session, such that employees respond either yes, I attended or no, I did not attend. In what follows, we include a sample of employee response data for illustration purposes, where each row contains a unique employees data and each column contains employees scores on each of the three attitudinal and behavioral concepts. To simplify things, the average of employees responses (i.e., scores) on the five-item measures for overall benefits satisfaction and turnover intentions have already been computed. Overall Benefits Satisfaction Turnover Intentions Attended a Benefits Information Session 3.78 2.87 Yes 4.60 1.91 Yes 3.19 2.14 Yes 4.12 1.90 Yes 3.88 2.90 Yes 3.84 1.64 Yes 4.68 1.63 Yes 3.46 3.29 Yes 3.26 2.45 Yes 4.52 2.07 Yes 2.06 2.98 No 2.84 2.97 No 3.63 3.01 No 3.36 3.17 No 3.64 3.45 No 2.84 4.09 No 2.71 3.40 No 2.84 3.14 No 2.86 2.83 No 2.93 2.66 No Given these data, we will attempt to answer the following questions: Is there a negative correlation between overall benefits satisfaction and turnover intentions, such that employees with higher overall satisfaction with benefits offerings have fewer intentions to leave the company? Do employees who attended a benefits information session have higher overall satisfaction with their benefits than employees who did not? To answer the first question, we can use simple linear regression, where overall benefits satisfaction is specified as the predictor variable and turnover intentions is specified as the outcome variable. Using Excel, we find the following: Coefficients Standard Error t Stat P-value Lower 95% Upper 95% Intercept 4.81 .61 7.89 .00 3.53 6.09 Overall Benefits Satisfaction -.60 .17 -3.49 .00 -.97 -.24 The results indicate that the regression coefficient for overall benefits satisfaction in relation to turnover intentions is .60, which means that the association between the two variables is negative. Such an association means that for every one-point increase in overall benefits satisfaction, we tend to see turnover intentions drop by .60 points. The corresponding p-value is less than the conventional two-tailed cutoff (alpha) value of .05, which means we can treat the regression coefficient of .60 as being statistically significant. Together, these two pieces of information provide evidence that, indeed, employees with higher overall satisfaction with benefits offerings have fewer intentions to leave the company. Regarding the second question, we can run an independent-samples t-test using Excel to determine whether the average overall benefits satisfaction score for those who attend a benefits information session is significantly higher than the average overall benefits satisfaction score for those who did not attend a session. Yes No Mean 3.93 2.97 Variance .29 .22 Observations 10 10 Pooled Variance .26 Hypothesized Mean Difference 0 df 18 t-Statistic 4.22 P(Tt) one-tail .00 t Critical one-tail 1.73 P(Tt) two-tail .00 t Critical two-tail 2.10 The results indicate the t-statistic that corresponds to the difference between the two means (averages) is 4.22, and the associated two-tailed p-value is less than the conventional cutoff of .05. Based on this information, we have evidence that in fact there is a statistically significant difference between the average overall benefits satisfaction score for those who attended a benefits information session and the average overall benefits satisfaction score for those who did not attend a session. To determine whether those who attended the information session had a higher average, we can look at the mean scores. The mean for the group of employees who indicated yes, I attended was 3.93, whereas the mean for those who indicated no, I did not attend was 2.97. Thus, we found support that indeed those who attended an information session tended to have higher satisfaction with the companys current benefits offerings. Excel Extension: Now You Try! On edge.sagepub.com/bauer, you will find an Excel exercise that provides additional practice evaluating employees satisfaction with benefits offerings. Using regression and independent-samples t-tests, you will test different hypotheses and answer different questions based on employee survey data