Interpreting Contingency Table Results
INTERPRETING CONTINGENCY TABLE RESULTS 6
Interpreting Contingency Table Results
[Student’s Name]
[Institution Name]
Interpreting Contingency Table Results
American jails are faced with numerous staffing problems such as the inability to recruit an adequate number of highly qualified staff and increased staff turnover (Stinchcomb & Leip, 2011 Stinchcomb & Leip, 2013). These problems are attributable to the extensive functions that jails have to undertake, coupled with inadequate funding. Consequently, jail staff have to contend with high-pressure work environments that negatively affect their morale. In this regard, this report is based on the results of the analyses performed using the National Jail Staff Data. The report will provide insights to the National Sheriff’s Association (NSA) and Sheriff Greg Champagne, the NSA president, on the levels of agency commitment and recommendations on measures to improve staff commitment (Curry, 2016).
There are three main sections in the report. The ensuing section presents an overview of the data, including the variables, hypotheses and descriptive statistics. Then, the results section presents the findings of inferential analysis using Chi-square. Finally, based on the findings, a recommendation section outlines the options that NSA should examine to enhance staff engagement.
Overview of the Data
Variables
For the current study, we utilize five main variables from the National Jail Staff Data. The variable for gender (male or female) is used as the intervention or control variable while the variable assessing commitment to agency (committed or not committed) is set as the dependent variable. Both gender and commitment to agency are nominal level measurements comprising two categories each. The independent variables include variables assessing whether the current pay is commensurate with the skills, education and experience of the jail staff, employee involvement in decision-making and whether the junior staff are accorded respect by the senior staff. All the independent variables are nominal level measurements with two categories each (yes or no).
Hypotheses
The study has three main hypotheses that assess the level of agency commitment among the employees. These are formulated as follows: (i) the staff compensation influences commitment to agency when controlling for gender, (ii) involvement in decision-making influences commitment to agency when controlling for gender, and (iii) respect for junior staff by the senior staff influences commitment to agency when controlling for gender.
As presented in Table 1, the sample of jail staff comprised 71.1% males and 28.9% females. Among the respondents, 91.3% indicated agency commitment while 8.7% indicated that they were not committed to the agency. About 69.5% of the respondents affirmed that they were being paid for what their skills, education and experiences are worth while 30.5% of the respondents did not agree. Additionally, 56% of the respondents indicate that they had a say in how things were done in their organizations while 44% indicated not having a say. Further, 74.3% of the respondents reported being respected by those above them in their chain of command while 25.7% reported not being respected by those above them.
Table 1. Descriptive Statistics for National Jail Survey
Name of Variable |
||
Commitment to Agency |
8.7% not committed |
91.3% committed |
Being paid what my skills, education and experience are worth |
30.5% no |
69.5% yes |
Having a say in how things are done here |
44% no |
56% yes |
Being respected by those above me in the chain of command |
25.7% no |
74.3% yes |
Gender |
71.1% male |
28.9% female |
Results
Influence of Staff Compensation
Table 2 presents the partial tables for commitment to agency by views about compensation for males and females. Among the male jail staff who indicated being paid for what their skills, education and experience are worth, 93.9% were more likely to be committed while 6.1% were more likely to indicate no commitment. Similarly, among the males who indicated not being paid for what their skills, education and experience are worth, 83.4% were more likely to be committed while 16.6% were more likely to report no commitment. There was a significant relationship between compensation and commitment to agency, 2(1, N = 1773) = 34.81, p < .001. The coefficient for Lambda indicates a 1.8% reduction in error when compensation is used to predict commitment to agency among males. However, the reduction is not significant, p = .41. Based on the significant values of Phi and Cramer’s V, there is a weak relationship between compensation and commitment to agency (Fletcher, 2014).
Among the females who reported being paid what their skills, education and experience are worth, 94.6% were more likely to be committed while 16.6% were less likely to be committed. Similarly, among the females who reported not being paid for what their skills, education and experience are worth, 87.7% are highly likely to be committed while 12.3% are less likely to be committed. The relationship between compensation and commitment to agency is significant, 2(1, N = 1773) = 7.48, p < .05. Based on the value of Lambda, there is a 0.5% reduction in error when compensation is used to predict commitment to agency. The values of Phi and Cramer’s V are significant, p’s < .05, which indicates a weak relationship between compensation and commitment to agency (Fletcher, 2014).
Table 2. Commitment to Agency by Staff Compensation, Controlling for Gender.
N = 1773 Source of data: National Jail Survey
Male |
||
Being paid what my skills, education and experience are worth |
||
Commitment to Agency |
No |
Yes |
Not committed |
16.6% |
6.1% |
Committed |
83.4% |
93.9% |
Total |
100% |
100% |
Chi-Square = 34.805 |
Phi = .166 |
|
Significance of Chi-Square = .000 |
Significance of Phi = .000 |
|
Lambda = .018 |
Cramer’s V = .166 |
|
Significance of Lambda = .405 |
Significance of Cramer’s V = .000 |
|
Female |
||
Being paid what my skills, education and experience are worth |
||
Commitment to Agency |
No |
Yes |
Not committed |
12.3% |
5.4% |
Committed |
87.7% |
94.6% |
Total |
100% |
100% |
Chi-Square = 7.476 |
Phi = .121 |
|
Significance of Chi-Square = .006 |
Significance of Phi = .006 |
|
Lambda = .005 |
Cramer’s V = .121 |
|
Significance of Lambda = .873 |
Significance of Cramer’s V = .006 |
Influence of Involvement in Decision-Making
Table 3 shows the partial tables for commitment to agency by involvement in decision-making for males and females. Among the males who reported having a say in how things are done in their work environment, 96.5% were more likely to be committed. Similarly, among the males who indicated having no say on how things are done, 82.6% were more likely to be committed. The relationship between involvement in decision-making and commitment to agency is significant, 2(1, N = 1758) = 69.35, p < .001. The value of Lambda indicates a 10.5% reduction in error when using involvement in decision-making to predict commitment to agency. The reduction is highly significant, p < .001. Additionally, the values of Phi and Cramer’s V are significant, which indicates a moderate relationship between commitment to agency and involvement in decision-making.
Among the females who reported having a say in how things are done, 98.1% were more likely to be committed. Among the females who reported not having a say on how things are done, 84.6% were highly likely to be committed. The relationship between commitment to agency and involvement in decision-making is significant, 2(1, N = 1758) = 30.49, p < .001. When involvement in decision-making is used to predict commitment to agency among females, the value of Lambda indicates a reduction in error by 11.2%, which is highly significant, p < .001. The values of Phi and Cramer’s V are highly significant therefore, there is a moderate association between commitment to agency and involvement in decision-making.
Table 3. Commitment to Agency by Involvement in Decision Making, Controlling for Gender.
N = 1758 Source of data: National Jail Survey
Male |
||
Having a say in how things are done here |
||
Commitment to Agency |
No |
Yes |
Not committed |
17.4% |
3.5% |
Committed |
82.6% |
96.5% |
Total |
100% |
100% |
Chi-Square = 69.354 |
Phi = .235 |
|
Significance of Chi-Square = .000 |
Significance of Phi = .000 |
|
Lambda = .105 |
Cramer’s V = .235 |
|
Significance of Lambda = .000 |
Significance of Cramer’s V = .000 |
|
Female |
||
Having a say in how things are done here |
||
Commitment to Agency |
No |
Yes |
Not committed |
15.4% |
1.9% |
Committed |
84.6% |
98.1% |
Total |
100% |
100% |
Chi-Square = 30.494 |
Phi = .247 |
|
Significance of Chi-Square = .000 |
Significance of Phi = .000 |
|
Lambda = .112 |
Cramer’s V = .247 |
|
Significance of Lambda = .000 |
Significance of Cramer’s V = .000 |
Influence of Respect for Junior Staff by the Senior Staff
Table 4 presents the partial tables for commitment to agency by respect for juniors by senior staff for males and females. Among the males who reported being respected by those above them in the chain of command, 92.7% were highly likely to be committed. Similarly, among the males who reported not being respected by those above them in the chain of command, 70% are more likely to be committed. The relationship between commitment to agency and respect for juniors by seniors is highly significant, 2(1, N = 1730) = 197.73, p < .001. Additionally, Lambda indicates a 15.4% reduction in error when using respect to predict agency commitment, which is a highly significant reduction, p < .001. The values of Phi and Cramer’s V are significant, which indicates a very strong association between commitment to agency and respect for juniors by senior staff.
Among the females who reported being respected by those above them in the chain of command, 97.7% were more likely to be committed. Among the females who indicated not being respected by those above them in the chain of command, 78.5% were more likely to be committed. The relationship between commitment to agency and respect for juniors by seniors is significant, 2(1, N = 1730) = 52.13, p < .001. The value of lambda indicates a 12.6% reduction in error when the respect for juniors by seniors is used to predict commitment to agency. Based on the significant values of Phi and Cramer’s V, there is a strong association between commitment to agency and respect for juniors by senior staff.
Table 4. Commitment to Agency by Respect for Juniors by Senior Staff, Controlling for Gender.
N = 1730 Source of data: National Jail Survey
Male |
||
Being respected by those above me in the chain of command |
||
Commitment to Agency |
No |
Yes |
Not committed |
30% |
2.8% |
Committed |
70% |
97.2% |
Total |
100% |
100% |
Chi-Square = 197.726 |
Phi = .40 |
|
Significance of Chi-Square = .000 |
Significance of Phi = .000 |
|
Lambda = .154 |
Cramer’s V = .40 |
|
Significance of Lambda = .000 |
Significance of Cramer’s V = .000 |
|
Female |
||
Being respected by those above me in the chain of command |
||
Commitment to Agency |
No |
Yes |
Not committed |
21.5% |
2.3% |
Committed |
78.5% |
97.7% |
Total |
100% |
100% |
Chi-Square = 52.125 |
Phi = .325 |
|
Significance of Chi-Square = .000 |
Significance of Phi = .000 |
|
Lambda = .126 |
Cramer’s V = .325 |
|
Significance of Lambda = .000 |
Significance of Cramer’s V = .000 |
Recommendations
In the first hypothesis, we posited that staff compensation influences commitment to agency when controlling for gender. The results support the hypothesis and indicate that high satisfaction with the levels of compensation results in more staff reporting commitment to agency. Importantly, the results are similar for both males and females. However, the weak association between the variables means that the level of precision is low when predicting the commitment to agency using the staffs’ sentiments on whether their salaries are commensurate with their skills, education and experiences. These results are indicative that pay does not have a substantial impact on engagement. Chamorro-Premuzic (2013) affirms the weak association between salary and job satisfaction. In this regard, it is important for NSA to understand the rewards that jail staff value the most to facilitate measures that enhance engagement.
Additionally, we hypothesized that involvement in decision-making affects commitment to agency when controlling for gender effects. The analysis findings support this hypothesis and indicate that increased involvement in decision-making leads to high levels of commitment to agency among male and female staff. The strong association between the variables indicates that having a say in how things are done is an important predictor of commitment to agency. These findings are consistent with those by Appelbaum et al. (2013) that lack of sufficient involvement of workers in decision-making lowers the level of satisfaction and commitment. In this regard, NSA needs to empower jail staff to make and account for their own decisions and outcomes.
Further, we hypothesized that respect for junior staff by their seniors would affect the commitment to agency when controlling for gender. The results indicate support for this hypothesis and show that both male and female staff are more likely to be committed when they feel respected by those above them in the chain of command. The variable for respect is a very important predictor of commitment to agency as shown by the strong association between the variables. According to Lipman (2017), respect enhances the moral of employees and motivates them to perform at their best. Therefore, all staff should be respected regardless of their job position to enhance commitment as well as productivity.
Conclusion
Ensuring commitment among jail staff is pertinent to curtailing the high staff turnover and improving the working environment. The findings of this study establish that staff commitment is highly influenced by the involvement of staff in decision-making and respect for juniors by those above them in the chain of command. As a result, the study recommends that NSA should empower jail staff to make their decisions and ensure that all staff are respected. Further, there is a need to evaluate the type of rewards that substantially influence commitment.
References
Appelbaum, S., Louis, D., Makarenko, D., Saluja, J., Meleshko, O., & Kulbashian, S. (2013). Participation in decision-making: A case study of job satisfaction and commitment (part three). Industrial and Commercial Training, 45(7), 412-419. http://dx.doi.org/10.1108/ict-09-2012-0049
Chamorro-Premuzic, T. (2013). Does money really affect motivation? A review of the research. Retrieved April 1, 2017, from https://hbr.org/2013/04/does-money-really-affect-motiv
Curry, P. (2016). Meet the 2016-2017 NSA President. Retrieved March 30, 2017, from https://www.sheriffs.org/blog/Meet-2016-2017-NSA-President
Fletcher, J. (2014). Crosstabulation with nominal variables. Retrieved April 1, 2017, from http://groups.chass.utoronto.ca/pol242/Labs/LM-3A/LM-3A_content.htm
Lipman, V. (2017). Why respect is key to employee engagement. Retrieved April 1, 2017, from https://www.forbes.com/sites/victorlipman/2017/01/10/why-respect-is-key-to-employee-engagement/#20e87ebd5394
Stinchcomb, J., & Leip, L. (2011). Recruiting the next generation of jail employees: Does generational relevance or agency size make a difference? American Journal of Criminal Justice, 37(3), 452-470. doi: 10.1007/s12103-011-9124-4
Stinchcomb, J., & Leip, L. (2013). Turning off jail turnover: Do generational differences matter? Criminal Justice Studies, 26(1), 67-83. doi: 10.1080/1478601X.2012.705785
No related posts.