Memorandum. Human Resources Division

Size: px

Start display at page:

Download "Memorandum. Human Resources Division"

Lambert Little
5 years ago
Views:

1 Memorandum Human Resources Division TO: FROM: RE: Vacellia Clark, Chief Examiner Civil Service Commission Human Resources Staff Establish a Passing Score for Animal Control Officer DATE: October 30, 2013 A. Summary City of Urbana Human Resources staff recommends a passing score of 50% using the application as the Civil Service Exam. This would result in an eligibility register of 12 candidates with no adverse or disparate impact. B. Background The position was open for applications from Sept. 20 Oct. 11, 2013 and Human Resources received 60 applications for the position. Numerically, the breakdown of applicants is as follows: Non-Minority 44 73% Minority 10 17% No response or n/a 6 10% Male 27 45% Female 29 48% No response or n/a 4 7% C. Application Screening The scoring plan utilized is explained in Appendix A of this memo. Required qualifications included at least two (2) years of relevant work experience, working knowledge of state laws and local ordinances, and working knowledge and experience with investigations. D. Passing Score At a passing score of 50 percent, adverse and/or disparate impact is not observed (additional data is attached). # % of Total % of Register Male % 50% Female % 50% No answer 0 Non-Min % 67% Minority % 17% No answer % 17%

2 Animal Control Officer Passing Score October 30, 2013 E. Attachments Appendix A: Application Exam Scoring Plan Appendix B: Disparate Impact Report for a 50% passing score Appendix A: Application Exam Scoring Plan 1. Education a. H.S. diploma/ged = 0 b. Some college = 0.5 c. Assoc. degree or higher (unrelated) = 1 d. Assoc. degree (related) = 2 e. Bach. Degree (related) = 3 f. Master's/higher (related) = 4 2. Experience: Type a. None = 0 b. Volunteer = 1 c. Combination of volunteer and professional = 2 d. Professional = 3 3. Experience: Scope a. Vet. Assistant = 1 b. Kennel /Shelter Attendant = 1 c. A closely related occupation/activity = 1 d. Animal control officer = 2 4. Experience: Amount a. Fewer than 2 years = 0 b. 2-4 years = 1 c. 5-8 years = 2 d years = 3 e. More than 15 years = 4 decisions you make have significant consequences) = 3 6. Experience with Investigations a. No experience = 0 b. 1-2 years' experience = 1 c. 3-5 years' experience = 2 d. More than 5 years' experience = 3 7. Experience with Breed Identification a. Yes = 1 b. No = 0 8. Certifications and Licenses a. One (1) point for every relevant certificate and/or license (e.g., veterinary technician, National Animal Control Association Academy, etc.). Total points possible = Knowledge of local and state laws/ordinances a. No experience = 0 b. Basic knowledge (infrequent exposure to laws and/or ordinances, and you do not make decisions based on these or decisions have little or no consequences) = 1 c. Moderate knowledge (occasional exposure to laws and/or ordinances; the decisions you make based on these laws/ordinances may have some consequences) = 2 d. Extensive knowledge (frequent exposure to laws and/or ordinances and the Page 2 of 2

3 Page 1 of 8 Disparate Impact Analysis (an On-Line Internet based application) Diversity Conference Instructions: Please fill out the information into the form below. Once you have entered your data below, you may select the types of analysis to be conducted by checking the appropriate boxes. Then press the compute button at the bottom of the form to view the results. Select the type of employment decision: Selection Enter a title for your report: Animal Control Officer 2013 Number of Male 27 Number of Non-Minority 44 Number of Younger Number of Non-Disabled 6 8 Number of Female 29 6 Number of Minority 10 2 Number of Older Number of Disabled -Adverse Impact -Chi-Square -Standard Deviation -Confidence Intervals Probability Distribution Select the Statistical Tests you wish to execute by checking or unchecking the boxes on the left. Then press the 'Compute' button below. Compute Display: Description of Statistic Interpretation of Results Animal Control Officer 2013 Adverse-Impact Report Adverse Impact and the "four-fifths rule." - A selection rate for any race, sex, or ethnic group which is less than four-fifths (4/5ths) (or eighty percent) of the rate for the group with the highest rate will generally be regarded by the Federal enforcement agencies as evidence of adverse impact. Uniform Guidelines on Employee Selection Procedures Rate of Females Rate of Males Adverse Impact Ratio for Females Adverse Impact Ratio for Males (6/ 29) = (6/ 27) = (0.2069/ )= 0.93 (0.2222/ )= 1.07 Adverse impact as defined by the 4/5ths rule was not found in the above data. Rate of Rate of Non- Adverse Impact Ratio for Adverse Impact Ratio for Non- (2/ 10) = 0.2 (8/ 44) = (0.2/ )= 1.1 (0.1818/ 0.2)= 0.91 Adverse impact as defined by the 4/5ths rule was not found in the above data.

4 Page 2 of 8 Chi-Square Report Observed Not Row Totals Expected 6 21 Males Females Column Total Chi-Square = The value of the statistic is less than This indicates that there is a 95 percent chance that these results have been obtained absent any form of bias. Therefore, you may conclude that these results fall within normal random variations and are not the result of bias. Observed Not Row Totals Expected 8 36 Non Column Total Chi-Square = The value of the statistic is less than This indicates that there is a 95 percent chance that these results have been obtained absent any form of bias. Therefore, you may conclude that these results fall within normal random variations and are not the result of bias. Standard-Deviation Report The difference between the proportion of the protected class and the proportion of all has a normal distribution with a mean and standard deviation. The statistic is shown below: (r / n) - p sqrt(p * (1-p) / n) * sqrt(1-q) Analysis of proportion of Females where: r = number of Females. n = number of (Females and Males). p = proportion of that are Females. Not Row Totals Males Females Column Total

5 Page 3 of 8 q = proportion of. r = 6 n = 12 p = 29 / 56 = q = (6 + 6) / ( ) = Standard Deviation Statistic = These results show that the proportion of Females is standard deviations below the proportion of. A result of less than 2 standard deviations is generally considered non-significant. Analysis of proportion of where: r = number of. n = number of ( and Non-). p = proportion of that are. q = proportion of. Not Row Totals Non Column Total r = 2 n = 10 p = 10 / 54 = q = (2 + 8) / ( ) = Standard Deviation Statistic = These results show that the proportion of is standard deviations above the proportion of. A result of less than 2 standard deviations is generally considered non-significant. Confidence Interval Report The proportion of the protected class has an expected value that would fall within a specified confidence interval. The statistic is shown below: Observed value = (r / n) Expected value = p Standard Deviation = sqrt(p * (1-p) / n) * sqrt(1-q) Confidence Interval: Lower Bound = p * Std Dev Upper Bound = p * Std Dev Analysis of proportion of Females where:

6 Page 4 of 8 r = number of Females. n = number of. p = proportion of Females among those. q = proportion of. r = 6 n = 12 p = (29/(29+27))=0.518 q = ((6 + 6)/( ))=0.214 (r/n)=6/12=0.5 The lower bound of the confidence interval is: (1.96* )= The upper bound of the confidence interval is: (1.96* )= Confidence Interval = to These results show that the proportion of Females Females (r/n=0.5) is contained in the confidence interval. Therefore a finding of disparate impact is not supported by this data. Analysis of proportion of where: r = number of. n = number of. p = proportion of among those. q = proportion of. r = 2 n = 10 p = (10/(10+44))=0.185 q = ((2 + 8)/( ))=0.185 (r/n)=2/10=0.2 The lower bound of the confidence interval is: (1.96* )= The upper bound of the confidence interval is: (1.96* )= Confidence Interval = to These results show that the proportion of (r/n=0.2) is contained in the confidence interval. Therefore a finding of disparate impact is not supported by this data. Probability Distribution Report Number Females Number Males Rate of Females Rate of Males Adverse Impact Ratio of Females Adverse Impact against Females? Cumulative Probability Probability 0 12 (0/29) (12/27) 0 YES (1/29) (11/27) YES

7 Page 5 of (2/29) (10/27) YES (3/29) (9/27) YES (4/29) (8/27) YES (5/29) (7/27) YES > 6 6 (6/29) (6/27) NO (7/29) (5/27) NO (8/29) (4/27) NO (9/29) (3/27) NO (10/29) (2/27) NO (11/29) (1/27) NO (12/29) (0/27) NO Given that 12 were from a pool of 27 Males and 29 Females it was possible to have from 0 to 12 Females. Adverse Impact would be found if you 5 or fewer Females. The probability of Adverse Impact occurring even if the employment decisions were random (i.e. unbiased) is (the sum of the probabilities of having 5 or fewer Females). Since the probability of Adverse Impact occurring even if the employment decisions were random (i.e. unbiased) is greater than 10%, an observed Adverse Impact may be not significant since the probability is greater than 1 in 10 that Adverse Impact would have occurred due to chance. Probability Distribution of the variable: Number of Females Number of female The probability distribution of having from 0 to 12 Females is displayed above. As can be seen, the most likely event (highest probability) to have occurred by chance (or decisions not affected by any form of bias) is to have 6 female. This represents the mean of the probability distribution. Approximately half of the probability distribution is above this point and approximately half is below this point. The total area contained in the probability distribution is equal to 1. Thus, probabilities for each number of female are a fraction of the total probability distribution. The larger areas of the distribution represent higher probabilities of occurance. Adding the individual

8 Page 6 of 8 probabilities up to a certain point enable you to compute the probability of having that many or fewer female. Adding the individual probabilities from a certain point and higher enable you to compute the probability of having that many or more female. The characteristics of the probability distribution--its mean and standard deviation--are a function of the number of female and male and the number of to be. Though it is possible to have from 0 to 12 female, the individual probabilities of having each number of female can be computed and accumulated. As noted before, these individual probabilities are a function of the number of female and male and the number of to be. Using the distribution above, a 90 percent confidence interval on the variable 'Number of Females ' would have a lower bound of 4 and an upper bound of 9. The significance of having 6 or fewer Females is graphically displayed below Number of female As noted earlier, Adverse Impact, according to the 4/5ths rule, would be found if you 5 or fewer female. You have 6 female. The probability of having 6 or fewer Females is equal to the cumulative probability for having 6 Females. The cumulative probability of having 6 female is and is graphically displayed, in red, above. Since the probability is greater than 10%, we are unable to reject the hypothesis that the decisions occurred due to chance. Therefore, we must conclude that it is entirely possible that having 6 or fewer female is an event that occurred due to chance and not from discriminatory actions by the employer. Number Number Non- Rate of Rate of Non- Adverse Impact Ratio of Adverse Impact against? Probability Cumulative Probability 0 10 (0/10) (10/44) 0 YES (1/10) (9/44) YES > 2 8 (2/10) (8/44) 1.1 NO (3/10) (7/44) NO

9 Page 7 of (4/10) (6/44) NO (5/10) (5/44) 4.4 NO (6/10) (4/44) 6.6 NO (7/10) (3/44) NO (8/10) (2/44) 17.6 NO (9/10) (1/44) 39.6 NO (10/10) (0/44) NO 0 1 Given that 10 were from a pool of 44 Non- and 10 it was possible to have from 0 to 10. Adverse Impact would be found if you 1 or fewer. The probability of Adverse Impact occurring even if the employment decisions were random (i.e. unbiased) is (the sum of the probabilities of having 1 or fewer ). Since the probability of Adverse Impact occurring even if the employment decisions were random (i.e. unbiased) is greater than 10%, an observed Adverse Impact may be not significant since the probability is greater than 1 in 10 that Adverse Impact would have occurred due to chance. Probability Distribution of the variable: Number of Number of minority The probability distribution of having from 0 to 10 is displayed above. As can be seen, the most likely event (highest probability) to have occurred by chance (or decisions not affected by any form of bias) is to have 2 minority. This represents the mean of the probability distribution. Approximately half of the probability distribution is above this point and approximately half is below this point. The total area contained in the probability distribution is equal to 1. Thus, probabilities for each number of minority are a fraction of the total probability distribution. The larger areas of the distribution represent higher probabilities of occurance. Adding the individual probabilities up to a certain point enable you to compute the probability of having that many or fewer minority. Adding the individual probabilities from a certain point and higher enable you to compute the probability of having that many or more minority.

10 Page 8 of 8 The characteristics of the probability distribution--its mean and standard deviation--are a function of the number of minority and non-minority and the number of to be. Though it is possible to have from 0 to 10 minority, the individual probabilities of having each number of minority can be computed and accumulated. As noted before, these individual probabilities are a function of the number of minority and non-minority and the number of to be. Using the distribution above, a 90 percent confidence interval on the variable 'Number of ' would have a lower bound of 0 and an upper bound of 4. The significance of having 2 or fewer is graphically displayed below Number of minority As noted earlier, Adverse Impact, according to the 4/5ths rule, would be found if you 1 or fewer minority. You have 2 minority. The probability of having 2 or fewer is equal to the cumulative probability for having 2. The cumulative probability of having 2 minority is and is graphically displayed, in red, above. Since the probability is greater than 10%, we are unable to reject the hypothesis that the decisions occurred due to chance. Therefore, we must conclude that it is entirely possible that having 2 or fewer minority is an event that occurred due to chance and not from discriminatory actions by the employer. View Source Code Copyright 1998, HR-Software.net All Rights Reserved. Send questions or comments to webmaster@hr-guide.com. Thank you.

Lecture Slides. Elementary Statistics Twelfth Edition. by Mario F. Triola. and the Triola Statistics Series. Section 7.4-1

Lecture Slides. Elementary Statistics Twelfth Edition. by Mario F. Triola. and the Triola Statistics Series. Section 7.4-1 Lecture Slides Elementary Statistics Twelfth Edition and the Triola Statistics Series by Mario F. Triola Section 7.4-1 Chapter 7 Estimates and Sample Sizes 7-1 Review and Preview 7- Estimating a Population