Data Research Analysis - Essay Example

? Due: PART II Introduction Research involves the application of a structured research process with an aim of answering certain questions or rather proving or disapproving a hypothesis. To answer these research questions or test whether a hypothesis may be approved or not, we employ a research design and method. Several procedures are involved including collecting data from the field however once data has been collected the next biggest challenge is to convert this data into useful information. This process (Data analysis) involves inspection, cleaning, transformation, and modeling of data with the goal of highlighting useful information so as to suggest conclusions, and support decision making (Box,G.E.P., Hunter, W.G., and Hunter, J.S. 1978). Several software have been developed (Hahn,G.J. and Meeker, W.Q. 1991), to help analyze data this include Statistical package for Social Sciences (SPSS), Stata, R, Minitab, SAS and many more. Data analysis in mixed methods research consists of analyzing the qualitative data using qualitative methods and the quantitative data using quantitative methods; therefore, knowing the steps in both forms of analysis is necessary in mixed research methods (Creswell & Plano Clark, 2007). For both forms of analysis, researchers go through the following similar set of steps: preparing the data for analysis, exploring the data, analyzing the data, representing the analysis, and validating the data. These procedures in quantitative research and qualitative research are different from each other as stated by Creswell and Plano Clark, (2007). In this project, to prove or rather to disapprove the hypothesis in an effort to answer the set of questions SPSS has been used to analyze the given set of data. The answers to the questions are thus given in the preceding paragraphs; Question A Hypothesis to be tested; a) H0: There is no significant difference in the salary earned by male and female employees b) H1: There is significant difference in the salary earned by male and female employees Since we have two groups (male and female) to compare their means and whose sample is less than 30 (a sample of 25 in this case), T-Test would be an appropriate test to use. T-Test is appropriate to be used to conduct the test since it assesses whether the means of two groups are statistically different from each other. From table 1 below, it can clearly be observed that the mean earning for the male employees is higher than that of the female employees. On average the male employees earn ? 24005 while the female employees earn 20319. Table 1: Report How much did you earn (?) last year Gender Mean N Std. Deviation Male 24005 13 3931.393 Female 20319 12 2370.797 Total 22236 25 3719.652 The difference can further be shown in the boxplot below; To answer part A of the questions i.e. whether there is any evidence that male and female employees differ in their salaries an independent samples test had to be conducted. From table 2 (Independent Samples Test), it can be observed that the p-value=0.010 (that is when the equal variances are assumed) which is less than the significance level (5%), we thus reject the null hypothesis and conclude that indeed there is a significant difference in the salaries earned by the male employees and the female employees. The male employees earn more than the female employees. Similar decision rule is made when unequal variances is assumed (p-value=0.10). Table 2: Independent Samples Test Levene's Test for Equality of Variances t-test for Equality of Means F Sig. t df Sig. (2-tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper How much did you earn (?) last year Equal variances assumed 2.710 .113 2.808E0 23 .010 3.686E3 1.313E3 970.142 6.401E3 Equal variances not assumed 2.863E0 1.994E1 .010 3.686E3 1.287E3 999.698 6.371E3 Question B Hypothesis to be tested; H0: There is no significant difference in the likelihood of raising a safety issue with the company’s H&S officer by male and female employees H1: There is no significant difference in the likelihood of raising a safety issue with the company’s H&S officer by male and female employees The mean rank score for the male employees regarding whether they have ever raised a safety issue with their company’s H & S officer is 13.73 while that of the female employees is 12.21 on a scale of 1 (Yes) and 2 (No) Table 3: Ranks Gender N Mean Rank Have you ever raised a safety issue with your company’s H&S officer Male 13 13.73 Female 12 12.21 Total 25 According to Anscombe, F. (1973), graphical representation of data is important since it gives a clear visual impression of the findings at a glance. From the chart below it can however be observed that there is a significant difference in the likelihood of raising a safety issue with the company’s H&S officer by male and female employees. Female respondents are more likely to raise safety issues compared to their male counterparts. For instance in the survey, 58% of the female employees raised safety issues with their Company’s H&S officer while only 46% of the male employees did the same. This gives a clear difference of about 12%. Question C Hypothesis to be tested; a) H0: There is no association between the period worked with this company and the salary earned by the employees b) H1: There is an association between the period worked with this company and the salary earned by the employees To test this, we had to check on the maximum and minimum salary and then convert the salary into ordinal i.e. 1 to represent ?17822-?21000, 2 to represent ?21001-?26000 and 3 to represent ?26001-?30561. Now, having this we used the chi-square test to test for the association between the two variables (salary earned and the period worked for the company). The table 5 below it can be seen that the highest paid employees had either worked for up to one year or between 1-3 years. However none of the employees who had worked for more than 3 years earned ?26001-?30561 which is the highest paid category. Table 5: Salary earned * How long have you worked with this company Cross tabulation Count How long have you worked with this company Total Up to 1 yr. 1-3 yrs. More than 3 yrs. salary ?17822-?21000 5 4 3 12 ?21001-?26000 2 3 5 10 ?26001-?30561 1 2 0 3 Total 8 9 8 25 Now using table 6 (Chi-square tests table) below, we can observe that there is no association (relationship) between the period worked for the company and the salary earned. 9 cells (100.0%) have expected count less than 5 and so we base our decision on Fisher’s exact test, the p-value for the Fisher’s Exact test is 0.521>0.05 (significance level) so we cannot reject the null hypothesis and hence we conclude that indeed there is no association between the period worked with this company and the salary earned by the employees. Table6 Chi-Square Tests Value df Asymp. Sig. (2-sided) Exact Sig. (2-sided) Exact Sig. (1-sided) Point Probability Pearson Chi-Square 3.866a 4 .424 .441 Likelihood Ratio 4.578 4 .333 .495 Fisher's Exact Test 3.579 .521 Linear-by-Linear Association .128b 1 .721 .861 .431 .131 N of Valid Cases 25 a. 9 cells (100.0%) have expected count less than 5. The minimum expected count is .96. b. The standardized statistic is .357. Question D QUESTION: Is there a relationship between employees’ answers to Q5 and Q6? Hypothesis to be tested; a) H0: There is no association between is there employees’ answers to whether the company is a safe place to work and H&S is always a top priority and whether employees enjoy their work and feel that they are valued member of staff b) H1: There is an association between is there employees’ answers to whether the company is a safe place to work and H&S is always a top priority and whether employees enjoy their work and feel that they are valued member of staff Explanation for the choice of test Since this question requires test of association/relationship it is important that we conduct a chi-square test. Testing for relationships is appropriate when two variables have been recorded for the same set of subjects or objects. The choice of test (in this case chi-square) is based on the fact that we are dealing with ordinal data. QUESTION: Do employees who have raised a H&S issue (Q1) differ in their views about working at the company as expressed in Q5 ? Hypothesis to be tested; H0: There is no significant difference in the employees view on whether the company is a safe place to work and H&S is always a top priority basing on whether the employees have ever raised a safety issue with your company’s H&S officer H1: There is significant difference in the employees view on whether the company is a safe place to work and H&S is always a top priority basing on whether the employees have ever raised a safety issue with your company’s H&S officer Explanation for the choice of test In this case we will use Kruskal-Wallis test which is a nonparametric test equivalent to the one-way ANOVA, which allows the comparison of more than two independent groups. Kruskal-Wallis is used when we wish to compare three or more sets of scores that come from different groups. The only disadvantage with this test is that it cannot tell you which specific groups were significantly different from each other; it only tells you that at least two groups were different QUESTION: Do employees who have raised a H&S issue (Q1) differ in their views about working at the company as expressed in Q6 ? Hypothesis to be tested; H0: There is no significant difference in the employees view on whether they enjoy their work and that they feel valued members of Staff basing on whether the employees have ever raised a safety issue with your company’s H&S officer H1: There is significant difference in the employees view on whether they enjoy their work and that they feel valued members of Staff basing on whether the employees have ever raised a safety issue with your company’s H&S officer Explanation for the choice of test Just like in the question tackled before, this question will apply the same test (Kruskal-Wallis test) to compare the differences in the feelings among the employees as to whether they enjoy their work and if they had raised issues before. REFERENCES 1. Anscombe, F. 1973, Graphs in Statistical Analysis, The American Statistician, PP.195-199. 2. Bernhardt, V. (1998). Data analysis for comprehensive schoolwide improvement. Larchmont, NY: Eye on Education, Inc. 3. Box,G.E.P., Hunter, W.G., and Hunter, J.S. (1978), Statistics for Experimenters: An introduction to Design, Data Analysis, and Model Building, John Wiley and Sons 4. Creswell & Plano Clark, 2007, Qualitative Research: Mixed research methods 5. Hahn,G.J. and Meeker, W.Q. (1991), Statistical Interval, John Wiley and Sons. Read More