Data Interpretation Practicum - Statistics Project Example

Data Interpretation Practicum Descriptive statistics, one way ANOVA, independent t test, and bivariate correlations were performed to determine if there was a relationship between several variables number of employees, safety behaviour of employees, injury rate, risks and safety climate according to various locations and gender of supervisor. The conclusion drawn from all statistical analysis was to accept the null hypothesis; whether there a relationship exists among variables. Introduction The purpose of this practicum is to examine the relationship between several variables at various work locations in Boston, Phoenix and Seattle. The variables identified included the number of employees, hours worked, gender of the supervisor, safety behavior of employees, injury rates, safety climate, and risk. The standard deviation, variance, and range are the descriptive statistics that were calculated using SPSS 16 statistics software. One way ANOVA, independent test and bivariate correlation are performed based on the research purposes. Research Question and Hypotheses Schindler and Cooper (2006) emphasizes the relationship among the variables are to be studied based on research questions to make prediction and policy making. The research question in this case is as follows: RQ: Is there any significant difference between various locations with respect to safety behaviour, injury rate, risk, safety climate etc. Do those variables differ with respect to gender of supervisor? The null and alternate hypotheses are as follows: H01: There is no significant difference in the variables between different work locations. H11: There is a significant difference in the variables between different work locations. H02: There is no significant difference in the variables between gender of supervisor. H12: There is a significant difference in the variables between gender of supervisor. Table 1 Descriptive Statistics of Variables that Impact Safety in the Workplace Descriptive Statistics N Range Mean Std. Deviation Variance Number of Employees 51 40 24.02 7.495 56.180 Hours worked 51 83200.00 49960.78 15590.23590 2.431E8 PerSafeBeh 51 .58 0.8655 .13891 .019 Injury Rate 51 76.92 15.1755 17.47447 305.357 Safety Climate 51 4.30 4.6971 1.03497 1.071 Risk 51 6 4.59 2.012 4.047 Descriptive Statistics Analysis & Assumptions The total number of employees surveyed over the two work locations was 51 (N). The work locations were Boston, (N= 15) and Phoenix (N=19) and Seattle (N=17). Employee behavior (PerSafeBeh) was defined as the percentage of acts that were deemed safe over a 12 month period. Injury rates were defined as the average number of injuries per 100 employees over a 12 month period. Safety climate was the employee’s perception of how their immediate supervisor prioritized safety over a 12 month period. Finally, risk was defined as the type of operational activity. In running the statistical analysis, the following assumptions were made: 1. The test variable has a normal distribution within the population. 2. A random sample from the dataset was utilized. 3. Test variable scores are independent. (Czitróm and Spagon, 1997) Table 2 ANOVA Results Descriptive Statistics Location wise Location No. of employees Mean Std. Deviation Std. Error Hours worked Boston 15 47008.0000 10482.86984 2706.66535 Phoenix 19 50357.8947 18753.08318 4302.25234 Seattle 17 52122.3529 15950.87753 3868.65605 Total 51 49960.7843 15590.23590 2183.06968 PerSafeBeh Boston 15 .7787 .17618 .04549 Phoenix 19 .9353 .08058 .01849 Seattle 17 .8641 .11397 .02764 Total 51 .8655 .13891 .01945 Injury Rate Boston 15 15.6293 13.87774 3.58322 Phoenix 19 17.1774 21.20512 4.86479 Seattle 17 12.5376 16.35585 3.96688 Total 51 15.1755 17.47447 2.44692 Safety Climate Boston 15 3.8527 .78951 .20385 Phoenix 19 5.3111 .97781 .22433 Seattle 17 4.7559 .77883 .18889 Total 51 4.6971 1.03497 .14493 Risk Boston 15 3.67 2.350 .607 Phoenix 19 4.89 1.823 .418 Seattle 17 5.06 1.713 .415 Total 51 4.59 2.012 .282 Table 3 ANOVA table Variables Source of variation Sum of Squares df Mean Square F Sig. Hours worked Between Groups 213210706.956 2 106605353.478 0.429 0.654 Within Groups 11939562061.672 48 248740876.285 Total 12152772768.627 50 PerSafeBeh Between Groups .206 2 .103 6.499 0.003 Within Groups .759 48 .016 Total .965 50 Injury Rate Between Groups 197.522 2 98.761 .315 0.732 Within Groups 15070.334 48 313.965 Total 15267.856 50 Safety Climate Between Groups 17.917 2 8.958 12.064 0.000 Within Groups 35.642 48 .743 Total 53.558 50 Risk Between Groups 18.289 2 9.144 2.385 0.103 Within Groups 184.064 48 3.835 Total 202.353 50 The decision had to be made whether to perform one-way ANOVA. The purpose of performing the one-way ANOVA is to analyze whether the means of a dependent variables were different among groups (Green & Salkind, 2011). For example, is the mean InjuryRate different by work location (Boston or Phoenix or Seattle)? In this case the variables Hours worked, PerSafeBeh, Injury Rate, Safety Climate and Risk were chosen for analysis, and one way analysis of variance test was best suited. From the one way ANOVA table, it is clear that there is a significant difference between the locations with respect to the variables safety behaviour of employees (probability 0.003), safety climate (probability 0.000). Other variables do not differ significantly viz. hours worked, injury rate and risk. Independent T test The independent t-test is performed to determine whether there exist significant difference between the gender of supervisor with respect to the variables hours worked, safety behaviour of employees, safety climate, injury rate and risk. 1. Each variable is bivariately normally distributed. 2. Random sample was obtained. 3. Test variables were independent. Table 4 Group Statistics Gender of Supervisor N Mean Std. Deviation Std. Error Mean Hours worked Male 24 53820.0000 12317.85978 2514.37260 Female 27 46530.3704 17527.79803 3373.22630 PerSafeBeh Male 24 .8304 .16502 .03368 Female 27 .8967 .10429 .02007 Injury Rate Male 24 13.5321 12.98843 2.65125 Female 27 16.6363 20.81543 4.00593 Safety Climate Male 24 4.3912 .86265 .17609 Female 27 4.9689 1.11292 .21418 Risk Male 24 4.79 2.000 .408 Female 27 4.41 2.043 .393 Table 5 Independent Samples Test Variables Levenes Test for Equality of Variances t-test for Equality of Means F Sig. T df Sig. (2-tailed) Mean Difference Std. Error Difference Hours worked .714 .402 1.698 49 0.096 7289.62963 4293.63441 1.733 46.642 0.090 7289.62963 4207.22299 PerSafeBeh 9.094 .004 -1.734 49 0.089 -.06625 .03821 -1.690 37.995 0.099 -.06625 .03921 Injury Rate 1.662 .203 -.629 49 0.532 -3.10421 4.93219 -.646 44.183 0.521 -3.10421 4.80381 Safety Climate 1.489 .228 -2.052 49 0.045 -.57764 .28145 -2.083 48.156 0.043 -.57764 .27727 Risk .052 .820 .677 49 0.501 .384 .567 .678 48.526 0.501 .384 .567 From the above table of independent test, it is observed that there is a significant difference between the gender of supervisor with respect to the variable safety climate (probability 0.045). Other variables do not show any significant difference. Table 6 Bivariate Correlation Results Number of Employees Hours worked PerSafeBeh Injury Rate Safety Climate Risk Number of Employees 1 1.000** -.021 -.636** .147 .351* .000 .885 .000 .303 .012 51 51 51 51 51 51 Hours worked 1.000** 1 -.021 -.636** .147 .351* .000 .885 .000 .303 .012 51 51 51 51 51 51 PerSafeBeh -.021 -.021 1 .035 .371** .089 .885 .885 .808 .007 .534 51 51 51 51 51 51 Injury Rate -.636** -.636** .035 1 -.013 -.433** .000 .000 .808 .930 .001 51 51 51 51 51 51 Safety Climate .147 .147 .371** -.013 1 .437** .303 .303 .007 .930 .001 51 51 51 51 51 51 Risk .351* .351* .089 -.433** .437** 1 .012 .012 .534 .001 .001 51 51 51 51 51 51 **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed). The Pearson correlation value of (-0.636** with probability 0.000) indicates that the number of employees is negatively correlated with injury rate. If more no. of employees are there, then the injury rate is less. Also there exists negative correlation between the variable injury rate and risk (-0.433 with probability 0.001) further supports the previous results of no relationship between injury rates and employee behavior at work. Positive correlation exists between risk and safety climate (0.437 with probability 0.001). The result shows that very meager non significant correlation exists between safety behavior of employees and injury rate (0.035 with probability 0.808). Opportunities for Further Study The present study based on several variables of work culture throws insight to the safety aspects of employees in their work places and provides the platform for the future studies regarding safety in the work culture. “Increasing evidence of employers and governments promoting and implementing policies and practices designed to reduce the horrendous toll of illness, injury, disease and death” (Bain 1997).The main limitation is being the sample size with 51 and if some more locations are included the study would throw more light regarding the safety aspects of employees in the work culture. A ‘safe’ organization—results from the constant engineering of diverse elements - for example, skills, materials, relations, and communications. (Gherhardi and Nicolini 2002). In additional to expanding the sample size, it is also recommended that the number of work locations be increased. Conclusion One way ANOVA, Descriptive statistics, independent t tests, and a Pearson correlation coefficient were performed to determine if a relationship existed between several variables involving different work locations and gender of supervisor. The important findings are 1. There is a significant difference between the locations with respect to the variables safety behaviour of employees (probability 0.003). 2. There is a significant difference between the locations with respect to the variables safety climate (probability 0.000). 3. There is a significant difference between the gender of supervisor with respect to the variable safety climate (probability 0.045) 4. The number of employees is highly negatively correlated with injury rate. 5. The result shows that very meager non significant correlation exists between safety behavior of employees and injury rate (0.035 with probability 0.808). References Bain, P. (1997). Human Resource Malpractice: the deregulation of health and safety at work in the USA & Britain, Industrial Relations Journal 28 (3): 176–191 Czitróm, V and Spagon, P. D. (1997). Statistical Case Studies for Industrial Process Improvement. SIAM. Gherhardi, S. and Nicolini, D. (2002). Learning the Trade: A Culture of Safety in Practice, Organization, 9 (2): 191-223 Schindler, P. S. and Cooper, D. R. (2006). Marketing Research. McGraw-Hill Education. Read More