EEOC Survey System Modernization Work Group Meeting March 8 - 9, 2012 DRAFT REPORT

Sage Computing, Inc.
March 19, 2012
Washington, DC

Table of Contents

Executive Summary

Background & Methodology

I. Day 1

II. Day 2

Appendix A-1: List of Participants

Appendix A-2: Contact with EEO-3 Respondents

Appendix B: Agenda

Appendix C: Breakout Group Instructions

Executive Summary

The Equal Employment Opportunity Commission (EEOC) organized a two-day Work Group Meeting on March 8th and 9th, 2012 at their headquarters building in Washington D.C. The objectives of this work group were to address:

• Modernization of the EEO-1, EEO-3 EEO-4, EEO-5 Survey systems to make data transfer more efficient.
• Improvement of EEOC reporting requirements and employer response/ process of EEOC requirements.
• Defining burden and steps to calculate the burden.

A total of 30 individuals representing survey methodologists, information technology and HRIS professionals, current survey respondents, vendors, and EEOC staff were invited to participate in the two-day meeting.

The two-day meeting was divided into several sessions. The participants spent day one on reviewing current survey design and its limitations; changes the respondents would like to see made to the existing systems; a discussion on ethnic and race categories; addition of reporting of pay data; and what impact these changes would have on the burden costs of respondents .Day two started with the respondents providing information on current costs. Two scenarios were presented to them and they were asked to provide an initial estimate of the costs of complying with the new requirements. The afternoon of day two saw the information technology, accounting and human resources experts, and the survey methodologists meet to review the recommendations from the respondents and provide recommendations to EEOC.

Recommendations provided at the end of the two-day meeting include:

Race/Ethnic Categories

• EEO-5 respondents would prefer EEO-1 race/ethnic categories.
• Provide option to not report ethnic category if employee refuses to self identify.
• Collecting and reporting multiple race/ethnicity of current employees would place an enormous burden on all survey respondents. The cost of re-surveying would be veryhigh.
• Reporting of multiple ethnic categories would require changes to the existing systems. This cost would also be very high especially for EEO-3, EEO-4, and EEO- 5 respondentsas current software vendors would make changes EEO-1 respondents could utilize but for other survey respondents, it would be custom changes. The cost of these softwareenhancements may be passed onto EEO-1 respondents also and is unknown at this time.
• Respondents would prefer they be given a list of 7 to 10 race/ethnic categories with "other" option to keep burden costs low.
• EEOC should consider removing/changing visual ID requirement.

• Prior to making changes to race/ethnicity categories EEOC should define the purpose of the requirements as these changes would have significant cost impacts.

Reporting of Pay Data

• EEOC pay data requirements should consider other reports such as Department of Labor.
• Reporting of means is easier than reporting pay bands.
• EEOC should specify the labor categories using the Department of Labor categories. This will ensure uniformity of reporting of salaries.
• EEOC should align reporting requirements with other agencies.
• EEO-1 respondents expressed concerns regarding confidentiality of data. Once the payroll information of a company was known, it would provide competitive advantage totheir competition. They were not in favor of reporting pay data.
• The cost burden of reporting pay data to EEOC would be minimal.

Modernization of Surveys

• Provide ability to upload files in multiple formats such as CSV and XML.
• Provide an Excel to XML conversion tool so respondents can use Excel to prepare the data.
• Provide an ability to upload data and be able to see errors instantly. The respondents should have "real time" access to fix/correct/edit the uploaded file.
• Allow the ability to include "The" in the respondent name.
• Ability to submit multiple files.
• EEO-1 survey should fix the NAICS and zip code files.
• EEOC should provide tutorial on their website (YouTube video as well as webinar).
• Standardize the definition of job type.

Overall, if the respondents are asked to report multiple race/ethnic categories, the cost burden on them would be very high. The highest cost would be to re-survey the existing employees andin procuring the software which captures multiple (as many that apply) races as identified by the employees. Keeping the reporting to a single field would be less costly as far as systemchanges are concerned but the re-survey and reporting costs would still be high.

Background & Methodology

The Equal Employment Opportunity Commission (EEOC) organized a two-day Work Group Meeting on March 8th and 9th, 2012 at their headquarters building. The objectives of this workgroup were:

• Modernization of the EEO-1, EEO-3 EEO-4, EEO-5 Survey systems to make data transfer more efficient
• Improvement of EEOC reporting requirements and employer response/ process of EEOC requirements
• Defining burden and steps to calculate the burden

The participants invited to be part of the meeting included:

• Survey respondents
• Vendors who complete survey for other companies
• Peoplesoft, SAP, and ADP experts
• Survey methodologists from two federal agencies
• Web technology/systems and database expert
• HRIS experts
• EEOC staff

A total of 30 individuals were invited to be part of the sessions. However, due to unforeseen circumstances, representatives from GE (EEO-1) and both EEO-4 respondents could not attend.In addition, there was no representation from EEO-3. A list of respondents invited to participate and their responses to the invitation are included as Appendix A.

The two-day meeting was divided into several sessions. The agenda for the working group is included as Appendix B. The participants were divided into two groups. The first group includedEEO-1 respondents and group two included EEO-4 and EEO-5 respondents. The information technology, human resources, and survey methodology experts were divided between the twogroups.

EEOC Commissioner, Mr. Stuart J. Ishimaru welcomed the participants and set the tone for the meeting by providing the need for adding multiple race/ethnic categories. This was followedby an address by Ron Edwards from the EEOC Office of Survey & Research Program. Mr. Edwards provided the goals for the two-day session and what EEOC expected to learn from theparticipants. He also provided the reason for modernization of the surveys.

After the joint session, the attendees broke off into break-out groups. The two groups were provided instructions (Appendix C) and asked to report back to the joint session at the end of thedetermined time frame.

The individual sessions were:

• Current Design of EEOC Surveys
• Discussion on Ethnic Categories
• Discussion on Collection of Pay Data
• IT Interfaces for EEOC Reporting Discussion
• Summary of Final Recommendations from the Respondents

The questions the participants were asked to discuss were:

• How the data are used?
• Relevance of collected information to organizations (both EEOC and stakeholder organizations?)
• How could submissions of data be streamlined?
• What are your operational procedures used to obtain the race and ethnicity of employees?
• Relevance of race and ethnicity to organizations (both EEOC and respondent organizations?)

• As multiple race and ethnicities become more prevalent, how can we collect that type of information and minimize record keeping and reporting burdens?
• Costs if new categories are required.
• Are there any current requirements that you report pay or earnings data?
• Discuss time/cost burden regarding each collection method.
• Could these requirements be adapted to reporting pay/earnings information to EEOC
• Are there other, more efficient methods to report pay/earnings-type information to EEOC?
• If reporting summary data, are there certain measures that are easier than others (means, medians, measures of variance, pay ranges)
• What types of confidentiality concerns exist and what safeguards would help alleviate these concerns?
• What are some of the current developments that EEOC could adopt that would reduce the burden of maintaining records and reporting to EEOC?
• What types of employer would and would not have access to these tools? How might small organizations be impacted?
• Can EEOC's processes be modified to reduce the burden?
• Are some industries different than others with respect to the use of IT in this area?

On day 2 of the meeting, the respondents spent the morning providing an estimate of their current and anticipated burden costs. The Work group participants were presented with a BurdenCost Data Table set up to capture the following key elements by each Survey Type (Companies (EEO1), State and Local Jurisdictions (EEO-4), and Schools (EEO-5):

• EEOC Survey Type.
• Job Type (position).
• Number of staff of this Job Type.
• Hourly Rate (average) for this Job Type.
• Other Cost Components (Fixed or Variable).

The data elements above were entered within the Burden Cost Data Table. The group was able to provide information for the EEO-1 and EEO-5 (school districts) data only during this session.

Once this information was entered the participants agreed to walk through and discuss how these same data elements may or may not be impacted by the following two (2) scenarios (thisdata is also captured in the Burden Cost Data Table):

• Scenario 1: The option of "check all that apply" is added to the corresponding survey.
• Scenario 2: Pay Data changes: Means/pay bands of hourly rates.

The data was collected and an average of the responses are included in this report (respondents expressed their desire to keep the actual data anonymous).

The afternoon session focused on a discussion between the information technology experts, the HRIS and survey specialists to provide recommendations to EEOC.

Details from each of the sessions are provided below.

I. Day 1

1.1 Welcome and Opening Remarks:
EEOC Commissioner, Mr. Stuart J. Ishimaru

Commissioner Ishimaru opened the meeting by emphasizing the importance of making the best of what is available, while creating proper metrics that retain the integrity of the datacollected. This all must be executed while maximizing the value of the taxpayers' dollar. He provided a background on the importance of collecting race and ethnicity data in the light of changing demographics.

Current challenges with the EEOC Survey Reporting Process:

• How do you collect data without destroying the integrity of it?
• Should people be able to check more than one box?
• Is discrimination occurring?
• How can we use statistics to see if discrimination is occurring?
• How can we use new IT technologies to help us meet our needs? More importantly, how do we use your (respondents) systems to meet joint needs?

1.2 Goals for the Work Group:
EEOC Director of Programs, Research & Survey Division, Mr. Ron Edwards

Mr. Ron Edwards provided the key goals of the workshop. A few key highlights from his discussion:

• EEOC surveys have remained very consistent over the years, (since 1965)
• About 43,000 employers filed EEO-1 reports in 1966. More than 67,000, employers reported by 2010.
• The private-sector workforce in 2010 was more than two and one third times its 1966 level from about 25 million to over 59.1 million.
• Women and ethnic diversity are increasing across our country.
• Increase in professional workers by 17%.
• 77% of workforce are high school graduates.
• 20% of workforce are college graduate.
• Inconsistencies in Race and Pay Data.
• EEOC is not taking advantage of the IT Systems available/used by respondents and may not be maximizing the potential of the off-the-shelf human resource and payroll packages.

These articulated goals transcend into the four (4) sessions (Questions and Brainstorming) planned for discussions on Day 1 (please reference Section 1.4 - 1.7 of this Output Report).

1.3. Participant Introductions/Process for Meeting/Breakout Group Instructions

Mr. Blevins asked each participant to introduce themselves and then outlined how the breakout groups would be divided:

• Group 1 will include EEO-1 respondents + IT + HR/Accounting + Survey Methodologists
• Group 2 will include EEO-4 and EEO-5 respondents + IT + HR/Accounting + Survey Methodologists

Instruction was then provided on what the breakout groups should be focused on (as well as conversation pieces to avoid), what roles (recorder, leader, etc) particiapants needed to play within each group, and general ground rules to uphold throughout the two-day meeting.

The ground rules for the meeting were as follows:

• Phones off or on vibrate
• Non-attribution
• Respect
• Open dialogue - honesty

Sections 1.4 through 1.7 of this Output Report capture the conversations that occurred during the Day 1 Breakout Group Sessions. Each session evaluated a specific topic and questions related to how EEOC will make decisions or adjustments to the surveys in the future

1.4. Session 1: Current Design of EEOC Surveys

Question 1 How is data used? Do you use for self-assessment, compliance and affirmative action, research?

Government: all of the above (self-assessment, compliance, affirmative action, research).

Large business/Small business: required to give report on small business contracts, affirmative action, diversity assessments and analysis. E.g.: EEO-1 report in PDF included as part of affirmative action plans.

Schools (Fairfax): affirmative action not included.

• Used as background checks/screening/etc for employment? NO
• Data is critical to NAICS codes.

Question 2 Relevance of collected information to organizations (both EEOC andstakeholder organizations)?

Do you ever have an issue (disability, older worker, etc.) that is not/cannot be addressed by the EEOC data?

• Age would be helpful from surveys and is difficult for corporations/organizations to collect.
• To reduce burden on employer with regard to duplicate data. Both VETS 100A and EEOC-1 request duplicative/similar data. EEOC does a better job of collecting this EEOC information.

Recommendation: EEOC and OFCCP to collaborate.
Recommendation: There needs to be one system versus two systems.

Question 3 How could submission of data be streamlined? How do we avoid missing information?

- Large Business: self identification requirement is difficult if hiring location/work location is different. Some organizations are now using a self-ID form to help with this. Cannot mandate self identification so visual identification is still an issue. EEOC requires visual identification at least.

Recommendation: EEOC to remove requirement for visual identification.

- Recruiting/hiring and on-boarding requirements on self-identification are different (not mandatory). This can cause errors with affirmative action numbers during reconciliation.

- Soft technology or other tools: new tool is cumbersome and not easily executed. (e.g.establishment IDs, zip codes, etc.).

Recommendation: Ability to upload, see errors, fix then submit to EEOC, should return detailed error messages and needed response. Have the ability to edit the text file and provide real time upload, purge, reload, etc.

1.5. Session 2: Ethnic Categories

Question 1 What are your operational procedures used to obtain the race and ethnicity of employees?

Group 1:

• Should be asked during on-boarding stage (some require a definite selection); not hiring stage.
• Some collect at applicant stage (usually they self-identify).
• Some companies use through on-line portal (intranet).
• Some companies outsource HR functions thus it becomes difficult to determine the race/ethnicity and some HR systems default to "white" if none is selected.

Group 2:

• Optional when they apply for the job/mandatory when hired.
• Can self identify first, and then use visual identification.

  Recommendation: Remove visual identification requirement.
  Recommendation: Add option to opt out.

Question 2 Relevance of race and ethnicity to organizations (both EEOC and respondent organizations)?

Group 1:

• Diversity.
• Benchmarking against own company or industry or geographical analysis.
• Request to see analysis of data from region, like size companies.
• Inconsistencies between data between EEOC data, affirmative action, etc.

Group 2:

• Diversity assessment.
• Compliance reporting.

Question 3 As multiple race and ethnicities become more prevalent how can we collect that type of information and minimize record keeping and reporting burdens?

Group 1:

• What is the PURPOSE of collecting additional races or ethnicities?
• Burden and costs for organizations to change /update this information in systems can be high.
• Match categories in EEOC, Census, OFCCP, etc.
• How often/should we resurvey? Every time a new category is added/changed the resurvey response may be low due to high cost and burden.
• If EEOC allows you to report on 2 or more races, how does EEOC plan to adjudicate at the end of the analysis and not show error?

Group 2:

• Align better with the EEO-1 categories. E.g. School districts are required to use the EEO-1 categories more.
• Include a new category titled, "multi-cultural" or like EEO-1 (2 or more).
• Give the option to opt-out versus organizational selection and subsequent ramifications.

Question 4 Costs to EEOC and/or Organizations if new categories are required?

Group 1:

• People categorized under old classification would be required to be re-classified and would place a large burden to organizations

Group 2 :

• People categorized under old classification would be required to be re-classified. This places a large burden to organizations and is a complex exercise. SAP, etc., do not currently allow for multiple boxes; but do allow for current single selection. Change would require customization of the SAP or other HRIS software to report accordingly.
• This would have a large impact on the integrity of the data that is being reported.
• EEOC should have a standardized approach that reduces burden to all organizations.

1.6. Session 3: Discussion on Collection of Pay Data

Question 1: Are there any current requirements that you report pay or earnings data?

Group 1: (combined question 1 and 3).

• Unemployment, compensation audits, DOL, other basis for benchmarking.
• Down to job title level or employee level.
• Unemployment DOL OFCCP match closely to EEOC reporting requirements for pay data.

Group 2:

• Information is public (EEO-5). There is no current consistent reporting requirement. Sometime newspapers ask for data and custom reports are provided.

Question 2: Discuss time/cost burden regarding each collection method.

Group 1:

• For IT, need someone to design modules for data collection
• Types of labor categories required to report vary.
• Upgrading HRIS system or other software.
• Training staff to report (product and report).
• Ability of company to spend money on a specific software tool or module impacts the cost/burden model
• Is the company using the correct software package? E.g. Peoplesoft and if so do they have the current version.
• Should it be done manually or through system?
• It is not easy to get to the underlying database to make changes and so will be quite burdensome to make any changes.

Group 2:

• Types of labor categories required to report vary.
• How will EEOC ensure that the right labor categories are being reported?
• There is no or little cost to report pay data but ethnic categories will be costly.

Question 3: Could these requirements be adapted to reporting pay/earnings information to EEOC?

Group 1:

Combined questions 1& 3

Group 2:

- N/A

Question 4: Are there other more efficient methods to report pay/earnings-type information to EEOC?           

Group 1:

• If subcontractors must submit like on EEO-4 form, then inefficient.
• Different salary ranges depending on geographic region.
• If must submit to OFCCP for audit purposes, could we share that data between OFCCP and EEOC?
• If every so many years the companies (based upon industry), were asked to report to EEOC, would that be easier? That really doesn't reduce burden due to the first-time setupcosts for a survey or modification to survey that is required.

Group 2:

• If rules were defined: don't ask for historical compensation data.
• Timing: keep it same as for EEO-5.

Question 5: If reporting summary data are there certain measures that are easier than others (means, medians, measures of variance, pay ranges)?

Group 1:

• If existing EEO-1 categories are used, then it would not be easier to use salary ranges.
• How is salary ranges calculated: overtime, bonus, overall compensation equalities?
• Hourly employees versus salaried employees (salary versus W2 data will vary by category and industry).
• MEDIAN (calculated field) is easier than asking for salary categories.

Group 2:

• Type of compensation needs to be defined clearly.
• Fringe benefits should be included (e.g. Not paid type fringe benefits).
• No issue with EEO-4 like reporting.
• Want to maintain the job categories from EEO-5
• Job classifications: is everyone calling the same "type" of job Analyst? How could we standardize this?
• Variations of regional, skill, market data: how does EEOC account for that?
• Most packages are designed for EEO-1 and EEO-4 and EEO-5 reports require additional customizations which is very costly.

Question 6: What types of confidentiality concerns and what safeguards would help alleviate these concerns?

Group 1:

• Privacy Act is a concern.
• If asking for aggregate data, then confidentiality is not a concern.
• Companies may not want to report data to a certain level due to competition reasons, etc.
• Statistical controls would be required to ensure the confidentiality of data to companies with smaller job categories.

Group 2:

• No privacy concerns because data is already public.
• If "cell" size lower than 5 we do not publish data for that cell.
• If decision to expand those categories is further decomposed and we have more cells, the more likely we are to run into this confidentially issue.

1.7. Session 4: IT Interfaces for EEOC Reporting

What are some of the current developments that EEOC could adopt that would reduce the burden of maintaining records and reporting to EEOC?

Group 1:

Methods Of Submission:

• Preferred method would be XML with instant validation checks.
• Smaller companies may like a tool to be made available (such as MS Excel conversion to XML format).
• CSV format.
• Web service like API to transmit the data versus direct upload to EEOC.

Staging Area:

• Would be able to see your upload, fix and then reload; could also load a few (more than one) file to see if data structure was correct for submission).
• Preload sample data.

Other Issues:

• Does EEOC monitor errors from audits of companies; perhaps EEOC should review these error messages.
• Zip codes: EEOC not matching with USPS.
• NAICS codes /tables need updating.
• Companies need the ability to update unit numbers
• Some addresses do not begin with a number; request that article "THE" in the field.
• Message center for alerts on the EEOC website to align with any messages or notifications from EEOC.
• If merger or acquisition goes through, an on-line EEOC form would be in place to allow the companies to depict the merged company unit numbers.

Group 2:

See summary of group 2 for all of Session 4 below.

What types of employer would and would not have access to these tools? How might small organizations be impacted?

Group 1:

• Paper employers want to do everything on paper due to spite, lack of trust, etc.
• Should be able to go online and do this; thus, XML may not be beneficial for a one/two person company. Good for small business.
• Tutorial on the site will help small businesses and new users.
• Tutorial on YouTube and/or EEOC Site would need to take a small quiz to be able to start using the EEOC forms.
• Include FAQ/troubleshooting/Lessons Learned Guide that is searchable on the EEOC site.

Group 2:

See summary of group 2 for all of Session 4 below.

Can EEOC's processes be modified to reduce the burden?

Group 1: n/a

Group 2: n/a

Are some industries different than others with respect to the use of IT in this area?

Group 1:

• Not really by industry; government contractors better suited to handle these needs.
• Larger companies may have issues, as they have more locations to report.
• Merger and acquisitions may impact difficulty.

Group 2:

**Note: in answering all questions for Session 4 (IT), Group 2 gave a list of general recommendations as per categories (highlighted below).

How We Are Doing This Today :

• On-line form.
• No HR systems supporting the EEO-5 format.
• Data file upload challenges: validation of data in the file; cannot see errors and correct until later (delayed).
• School system that report for all school districts: all districts have to verify before submission and so delay in process and time.

How We Would Like It To Work Better:

• Multiple businesses and packages: file upload (with browse ability to locate and select file to upload); multiple formats (Excel, XML, and other formats depending on feasibility of varying organizations.
• Have staging area: data before submissions can be reviewed, errors corrected (edits) and validated (type of data; fields, etc.).
• If major revamp: can we replace the old file and replace with new updated file.
• Multiple locations: ability to submit multiple files to EEOC versus summary of all locations.

Wish List.

• Allow companies with multiple business units to have individual business units submit and validate their respective reports prior to overall submission.
• Audio training. Placed on the EEOC Website prior to being able to upload and utilize the on-line form.
• YouTube Video
• Make all the changes that are approved once versus various changes over multiple time frames.
• Multiple file format types: send out a survey to see what type of formats would be preferred by users.
• Break out the error by entity…if 12 entities submitted, we can specify that entity 5 needs correction.
• On-line validation will help reduce the burden and lag time.

1.8. Closing Remarks, Summary, Next Steps for Day 2

At the end of Day 1, participants were asked to complete a "one-word checkout" summarizing their thoughts on the session thus far. The following words were chosen to describe the progress:

• Progressive
• Perspective
• Productive
• Open
• Hopeful
• Appropriate
• Informative
• Engaging
• Effective
• Positive
• Pleased
• Collaborative
• Insightful
• Educational
• Helpful

 

II. Day 2

2.1. Burden Cost Data Attributes & Scenario Discussions

For Day 2 Burden Cost Data discussions, the Work group participants were presented with a Burden Cost Data Table set up to capture the following key elements by each Survey Type (Companies (EEO-1), State and Local Jurisdictions (EEO-4), and Schools (EEO-5):

• EEOC Survey Type.
• Job Type (position).
• Number of staff of this Job Type.
• Hourly Rate (average) for this Job Type.

• Other Cost Components (Fixed or Variable).

The data elements above were entered within the Burden Cost Data Table. The group was able to provide information for the EEO-1 and EEO-5 (Schools) data only during this session.

Once this information was entered, the participants agreed to walk through and discuss how these same data elements may or may not be impacted by the following two (2) scenarios (this data is also captured in the Burden Cost Data Table):

• Scenario 1: The option of "check all that apply" is added to the corresponding survey.
• Scenario 2: Pay Data changes: MEANS of hourly rates.

Table 1 provides a summary of the data provided by the participants. The numbers represent an average of the estimates provided by the participants. The rates and the hours reported by the participants were multiplied to get the labor costs. Any additional charges were added to this estimated cost. The average cost by survey type was then calculated. The EEO-1 numbers are based on a total of five (5) responses.

The additional costs reported in Table 1 are the average additional charges that participants thought would be incurred if EEOC required reporting of race/ethnicity and pay data. For example, $1090.27 is the average additional expense based on responses provided by all EEO-1 participants to provide pay data.

It should be noted that all participants in the Work Group Meeting currently respond by uploading data files. If completing the survey by entering the data online, the costs would be much higher to report pay data.

 

Meeting participants provided an estimate of the increase in labor hours for scenario 1 and 2. Table 2 provides the estimated number of additional hours by labor category for all surveys.These are the additional staff hours that would be required if EEOC implemented both the scenarios.

The variable costs reported to comply with the new reporting requirements could not be quantified. The respondents reported that the cost of re-surveying the workforce would be incredibly high. The school districts said the costs are dependent on school budgets and they may not be able to get the money to re-survey. They were unwilling to specify the dollar amount. It was also reported that it would be virtually impossible to do a visual identification of the existing employees for those with multiple locations, workers who work from home, and other such factors. The other one-time costs also would include the cost of getting new modules/updates for their human resources and accounting packages. Since all packages provide reports only to comply with EEO-1 reporting, the costs to EEO-3, EEO-4, and EEO-5 would be much higher as it would be customization of the software.

2.2. PayOff Matrix

The objective of the PayOff Matrix exercise is to pull information together from Day 1 into a summarized list (1-n) of recommendations for survey design and systems improvement (storyboarding) using the PayOff Matrix methodology.

The Payoff Matrix is a two-by-two table that helps sort and evaluate ideas. Typical dimensions include: Payoff vs. Ease of Change; Level of Difficulty vs. Benefit; Customer Impact vs. Organizational Impact; Required Resources vs. Time Required to Implement; Global vs. Local; etc. The primary reason to execute a payoff matrix is to sort and prioritize ideas/actions;serving as a dimensional tool to take a range of ideas (1-n) and narrow down to how to best spend the group's (EEOCs) time and energy moving forward. The participants and EEOC agreed to use the Payoff Matrix dimensions as found in Figure 1.

Figure 1: Payoff Matrix Template

*note: Stakeholders includes both the EEOC and Respondents.

Those actionable 1-n items in QUAD 1 of the Payoff Matrix in Figure 2 are the primary target actions for EEOC to consider and execute. It is recommended that these Quad 1 (High Benefit, Lease Time required) actions be further prioritized by EEOC and a project/actions execution schedule be developed accordingly.

Figure 2: EEOC Survey System Modernization Work Group Meeting Payoff Matrix.

Figure 3: EEOC Survey System Modernization Work Group Meeting Payoff Matrix Key.

Figure 4: EEOC Staff Facilitates the PayOff Matrix Exercise

At the conclusion of the PayOff Matrix exercise, the IT System participants and the Survey Methodologist participants were asked to stay for the final afternoon session and discussions. If available, all other participants were welcome to stay as well.

2.3 Burden Cost Feasibility and Additional Cost Considerations

The final session/discussion focused on the feasibility of Scenario 1 or Scenario 2 (reference Section 2.1 of this report for further clarification on scenarios) from the perspective of the IT Systems participants and the Survey Methodologists.

Scenario 1 comments:

The option of "check all that apply" is added to the corresponding survey.

Recommendation A:

• List all categories (5 to 7) and then have the option of YES/NO for each field.
• EEOC should not want to list 50-60 options with the YES/NO field option (how could we indeed list all possible ethnic scenarios?).
• Do not want to use the ACS (American Community Survey) form as it has issues

Recommendation B:

• List as many possible ethnic scenarios as feasible.
• How many combinations should the EEOC have? Difficult decision.
• Re-surveying by respondents within their organizations will be time consuming/costly and respondents may not do it. The data collected by EEOC will not be accurate for the first few years.
• If EEOC lists many ethnic scenarios, may still leave someone off or may not answer due to confusion of options.

Recommendation C:

Use "other" category on survey because "other" may provide better data quality than bad or incorrect data.

SCENARIO 2 comments:

Pay Data changes .

EEOC informed participants that the National Academy of Science is currently researching how we should best collect pay data.

• Statisticians like pay bands, not means or medians.
• Respondents would report whatever it was that EEOC asked for but would prefer clear guidelines.
• EEOC needs to define the job types better and need to be cognizant of the fact that there are different pay bands by industries and job types/categories.
• Private employers are always going to be hesitant to report pay data; therefore, we need to aggregate data as best we can

2.4. Closing Remarks

At the end of Day 2, participants were asked to complete a "one-word checkout"; again summarizing their thoughts on the session. The following words were chosen to describe how the remaining participants felt as they departed on Day 2 and ended the Survey Methodology Workgroup meeting:

• Burden
• Value-added
• Informative
• Accomplished
• Productive
• Collaborative
• Voluntary
• Brevity
• Intensity
• Tired
• Scary
• Unbelievable
• Relieved

Appendix A-1: List of Participants

 

LIST OF WORKING GROUP MEMBERS REDACTED

 

*Note: Those highlighted in yellow were confirmed participants but did not attend at the last minute.

 

Appendix A-2: Contact with EEO-3 Respondents

 

LIST OF WORKING EEO-3 RESPONDENTS REDACTED

 

 

 

 

Appendix B: Agenda

 

 

Appendix C: Breakout Group Instructions

Breakout Guidance for EEOC Survey System Modernization Work Group Meeting Breakout Sessions

The following information is intended as guidance for breakout groups. The workshop facilitators will spend time with each group to provide any additional assistance necessary.

Group Focus

As employees' (users) familiar with the Equal Employment Opportunity Commission's (EEOC) survey and data collection systems, your focus is on evaluating specific topics related to these systems, and making suggestions and comments that will help the EEOC in developing its final set of recommendations for modernization. The group should avoid discussion on any non- survey/data related EEOC practices or policies.

Each group of users will participate in 4 separate breakout sessions on Day 1, lasting approx. 45 minutes each.

Member Roles

Group Recorder: The recorder will take notes based on the group's input, and then review key points with the group in the last 5-10 minutes of each session. The recorder will not evaluate or edit the group's findings, except for brevity.

Group Leader: The group leader will help guide the discussion along and ensure that the group stays on topic. Although he/she can answer general questions and clarify the intent of written materials, he/she will not provide opinions, nor exert personal/professional judgment on the group's findings. As the recorder provides a recap of the session's findings, the group leader will capture key points on the flipchart, and report the group's findings to the overall audience.

Group Timekeeper: The group timekeeper will ensure all participants are notified of the time and minutes remaining to ensure adequate and effective discussion. (e.g. 15 minute warning, 10 minute warning, 5 minutes warning, etc.).

Group Members: Individual group members will provide expert opinion to each of the stated questions.

Group Procedures

Step 1: Evaluate Session Questions (35 minutes)

The group should address as many questions as possible, with each group starting with the question number corresponding to its group number.

Your group does not necessarily need to come to consensus on a particular issue, and the recorder should ensure that all input is captured. Additionally, your group may not be able to address all of the questions in a particular session, due to time limitations. The workshop facilitators will note in the final report which questions could not be addressed, to ensure proper weighting is given to overall group findings.

Step 2: Group Review (10 minutes)

The recorder gives a summary of group's findings and the group provides any additional information that may have been missed. The group leader will transcribe key points to the flipchart (Butcher Block paper).

Step 3: Debrief to Larger Group (10-12 minutes each group).

Each group will have approximately 10 minutes to report back their discussions/answers to the larger group.