DCI STAFF PRESENT AT SIOP’S 26th ANNUAL CONFERENCE HELD IN CHICAGO

by David Morgan, Keli Wilson, Jana Moberg, Amanda Shapiro & Eric Dunleavy, Ph.D., DCI Consulting Group

The 26th Annual Conference for the Society for Industrial and Organizational Psychology (SIOP) was held April 14-16, 2011 in Chicago, IL. Attendees and presenters from a variety of industries and backgrounds took part in the event. As is typically the case for the Annual Conference, many sessions covered topics related to current EEO-related issues, as well as important themes concerning the workplace, such as assessment and selection, performance appraisals, adverse impact measurement, etc. DCI Consulting Group staff members were involved in a number of those presentations. Some conference presentation highlights are found below.

Recommendations of the Technical Advisory Committee on Adverse Impact Analysis

A respected panel of presenters including David Cohen (Co-chair), Eric Dunleavy (Co-chair), and Mike Aamodt from DCI Consulting; Mary Baker, ERS Group; John Geier, Paul, Hastings, Janofsky & Walker LLP; Lorin Mueller, American Institutes for Research; Mickey Silberman, Jackson Lewis LLP; and Evan Sinar, DDI, discussed some important recommendations of a 70-expert Technical Advisory Committee (TAC) on Adverse Impact Analysis, created by the Center for Corporate Equality (CCE). The session was structured around 15 critical questions addressed in the TAC, with important implications for contemporary measurement of adverse impact. The panel members did note that the results and recommendations were presented in the aggregate, and no single TAC member could be explicitly linked to a particular opinion, perspective or recommendation—survey results were anonymous, and specific focus group participants and comments were confidential. Thus, individual TAC members may have disagreed with some of the recommendations made.

The discussion included, but was not limited to, the following topics:

• Who is an “applicant” for the purposes of conducting an adverse impact analysis? Panel members noted three criteria to be considered an applicant, which include 1) showing interest in an open position, 2) properly following an organization’s rules for applying, and 3) meeting the basic qualifications; however, those applicants not actually considered for a position or those who formally withdraw from the process should generally not be included in the analysis.


• Should organizations “guess” the gender or race of applicants who do not self-identify their race or gender? The general TAC recommendation was that organizations should not “guess” the race or gender of applicants who do not self-identify. Although organizations are not legally required to backfill the race or gender of applicants who are hired, doing so is a reasonable practice.


• Which statistical significance tests should be used? Panelists noted that it is important to determine the context in which the data are being analyzed. The statistical model that best mirrors the reality of the employment decision process should be used; however, it is often difficult to understand how employment decisions were made, and thus challenging to identify the most appropriate model.

Adverse Impact Analysis: Contemporary Perspectives and Practices

Presenters in this panel offered modern perspectives on adverse impact analysis, including: statistical significance tests that currently dominate the EEO landscape and which groups to compare in such tests, practical significance tests (including an alternative to the 4/5ths rule), a comparison of three ‘multiple events’ tests for aggregated data, the impact of adverse impact from response distortion (faking) on assessments, and adverse impact-validity tradeoff scenarios.

Some interesting and important take-home points from the session included:

• A review of some non-traditional EEO enforcement activity by Federal agencies in recent years, in which Eric Dunleavy, DCI Consulting, noted that some recent OFCCP settlements placed emphasis on Non-Hispanics as the disadvantaged group. The importance of proactive analyses in which both traditional (Total Minority vs. White) and individual highest selected group analyses for bottom line impact was emphasized.


• An alternative to the 4/5ths rule, in which Seydahmet Ercan (with Frederick L. Oswald), Rice University, introduced a novel statistic that adjusts the selection ratio to improve upon a statistic that reflects already transformed ratios. The rule was referred to the lnadj rule.


• A finding presented by Elizabeth Howard (with Scott Morris), IIT, that although Multiple Events Tests provide an effective way to analyze aggregated data across samples, the outcome may differ depending on which test is used. The uncorrected Mantel-Haenszel test was found to have the best performance among the tests examined, with the highest power to detect a significant when such a difference occurred.


• A discussion by Phillip Mangos (with John Morrison), Kronos Inc., in which the results of a study indicated that high levels of focal group faking on employment assessments may reduce or reverse patterns of adverse impact.


• A review of the adverse impact-test validity tradeoff by David B. Schmidt (with Alexander R. Schwall and Evan Sinar), DDI, accompanied by a discussion on the potential over-focus on adverse impact statistics, emphasized that over-focus of such statistics may affect other important organizational goals (e.g., diversity).

Overhauling Hiring Methodologies: Unproctored, Automated Assessment in Federal Hiring Reform

Jone Papinchock of U.S. Office of Personnel Management, Ryan Shaemus O’Leary of Personnel Decisions Research Institutes, Laurie E. Wasko of HumRRO, Elaine D. Pulakos (Discussant) of Personnel Decisions Research Institutes & Brian S. O’Leary (Discussant) of U.S. Office of Personnel Management shared the development and implementation process of a nationwide testing program using unproctored, automated assessments in response to a memorandum signed by President Obama in May 2010. The symposium discussion focused on the following:

• History of testing (e.g., beginning with the pre-1883 Spoils System);


• Nationwide testing program goals (e.g., decrease applicant burden and improve quality and speed of hiring, capture assessment data to use and reuse, and allow job seekers to take assessments anytime and anywhere);


• Determination of jobs to include in initial testing program (e.g., 12 occupations that make up approximately 33% volume of jobs);


• Process of selecting the appropriate online, unproctored assessments, and;


• an overview of challenges and achievements encountered during the development process (e.g., controlling size and scope of project, buy in, successfully implementing the testing program for a handful of the 12 jobs).

The panel concluded with an overview of challenges and achievements encountered during the development process (e.g., controlling size and scope of project, buy in, successfully implementing the testing program for a handful of the 12 jobs).

OFCCP/Legal Defensibility Safeguards: Hit ‘em With Your Best Shot

Lilly Lin of Development Dimensions International led a symposium with David Schmidt of Development Dimensions International, Laura Mastrangelo Eigel of Frito-Lay North America, David Cohen of DCI Consulting Group, and Kevin Murphy of Pennsylvania State University on OFCCP trends and best practices in proactively dealing with regulatory agencies. The symposium focused on the following:

• Emerging trends seen in OFCCP audits over the previous year and legal updates, including the updates to the Americans with Disabilities Act Amendments Act (ADAAA) and proposed updates to the Vietnam Era Veterans' Readjustment Assistance Act (VEVRAA)


• Best practices for dealing with regulatory agencies and in conducting job analyses and collecting assessment validity evidence


• A case example of Frito-Lay’s effective proactive actions taken during a recent OFCCP audit

Abolish the Uniform Guidelines

Another interesting session focused on the current state and use of the Uniform Guidelines on Employee Selection Procedures (UGESP). UGESP are the federal regulations that inform on adverse impact measurement and, in cases where substantial adverse impact exists, on validation research strategies that are acceptable for demonstrating job-relatedness and/or business necessity. The esteemed panel included (1) Dr. Mike McDaniel, Virginia Commonwealth University, (2) Dr. Jim Sharf, Employment Risk Advisors, (3) Dr. James Outtz, Outtz and Associates, Dr. Art Gutman, Florida Institute of Technology, and (5) David Copus, Ogletree Deakins.

As many readers of this blog are aware, the Uniform Guidelines were written in 1978 and have not been revised. Some of the panelists emphasized that the UGESP were (1) ‘stale’, (2) inconsistent with contemporary research findings on adverse impact and job relatedness, and (3) uneven related to guidance from professional authorities like the AERA/APA/NCME Standards and the SIOP Principles. Given this perspective, these panelists suggested that UGESP should be revised or even abolished.

Other panelists suggested that (1) the UGESP provide some very useful guidance, (2) we should try not to ‘throw the baby out with the bathwater’, (3) courts have essentially updated and corrected UGESP in their interpretations and rulings and (4) it would be extremely difficult to revise the UGESP and toss out 33 years of relevant case law without any clear understanding of what would replace them. Given this perspective, these panelists seemed to suggest that the UGESP can still be useful in situations where they are applied and interpreted responsibly.

We suggest that readers interested in this topic also review a recent article found here: http://siop.org/article_view.aspx?article=850. The article, written by Mike McDaniel and colleagues at VCU in the journal Industrial and Organizational Psychology: Perspectives on Science and Practice, is titled The Uniform Guidelines Are a Detriment to the Field of Personnel Selection. Responses to this article are due by May 23rd, and will be published as a set later in the summer. We expect the set of response articles to be very diverse in perspective. Stay tuned.

On a closing note, some DCI staff members were formally recognized by SIOP at this year’s conference. First, Eric Dunleavy received the first SIOP Award for Distinguished Early Career Contributions-Practice. This award is given to a recipient who has made distinguished contributions to the practice of I-O psychology within seven (7) years of receiving a doctorate. Second, Art Gutman, who contributes to this blog spot and is a Professor at Florida Institute of Technology, was awarded SIOP Fellowship. Fellow status in SIOP is an honor granted through a nomination process. Society Fellows are distinguished industrial and organizational psychologists who have shown evidence of unusual and outstanding contributions or performance in I-O psychology, through research, practice, teaching, administration, and/or professional service. Congratulations to Eric and Art.

OFCCP OFFICIALLY ANNOUNCES NOTICE OF PROPOSED RULEMAKING (NPRM) ON VETERANS' AAP REGULATIONS

by David Cohen, Eric Dunleavy, Ph.D. and David Morgan, DCI Consulting Group

Yesterday OFCCP formally announced a proposed rule to strengthen affirmative action requirements for protected veterans. The announcement can be found below and here:

News Release

OFCCP News Release: [04/25/2011]
Contact Name: Michael Volpe or Gloria Della
Phone Number: (202) 693-3984 or x8666
Release Number: 11-0567-NAT

US Labor Department proposes rule to improve employment of protected veterans

WASHINGTON — The U.S. Department of Labor's Office of Federal Contract Compliance Programs today announced a proposed rule to strengthen affirmative action requirements of federal contractors and subcontractors for veterans protected under the Vietnam Era Veterans' Readjustment Assistance Act of 1974. Veterans protected by VEVRAA include those with disabilities and those recently discharged as well as those who served during a war, campaign or expedition for which a campaign badge is authorized. The proposed rule will be published in the April 26 edition of the Federal Register.

"At the Labor Department, we support veterans as they seek meaningful ways to apply their talents to expand the American economy. By re-examining our affirmative action requirements, we will ensure that our nation's veterans are protected against discrimination and provided equal opportunity in the workforce," said OFCCP Director Patricia A. Shiu.

The award of a federal contract comes with a number of responsibilities. Among them are complying with non-discrimination and affirmative action provisions, engaging in meaningful and effective efforts to recruit and employ veterans protected under VEVRAA, and maintaining accurate records on affirmative action efforts. Failure to abide by these responsibilities may result in various sanctions, from withholding progress payments to termination of existing contracts and debarment from receiving future ones.

The framework articulating a contractor's responsibilities with respect to affirmative action, recruitment and placement has remained unchanged since 1976. Increasing numbers of veterans are returning from tours of duty, and many are faced with substantial obstacles in finding employment upon leaving the service.

The proposed rule clarifies mandatory job listing requirements, under which a contractor must provide job vacancy and contact information for each of its locations to an appropriate employment service delivery system. The rule proposes requiring contractors to engage in at least three specified types of outreach and recruitment efforts each year. In addition, the proposed rule would require that all applicants be invited to self-identify as a "protected veteran" before they are offered a job. Increasing data collection on job referrals, applicants and hires, and requiring contractors to establish hiring benchmarks to assist in measuring the effectiveness of their affirmative action efforts also are proposed.

Comments on the notice of proposed rulemaking must be submitted by June 25. Visit the federal e-rulemaking portal, http://www.regulations.gov/, to submit comments.

In addition to VEVRAA, OFCCP's legal authority exists under Executive Order 11246 and Section 503 of the Rehabilitation Act of 1973. As amended, these three laws hold those who do business with the federal government, both contractors and subcontractors, to the fair and reasonable standard that they not discriminate in employment on the basis of gender, race, color, religion, national origin, disability or status as a protected veteran. For general information, call OFCCP's toll-free helpline at 800-397-6251. Additional information is available at http://www.dol.gov/ofccp.

The proposed rule had been approved by OMB earlier in the month, and will be published in today’s Federal Register. Importantly, the new rule includes sections related to veteran outreach and recruitment efforts, as well as to the use of quantitative metrics to assess the success of those efforts. Comments on the notice of proposed rulemaking must be submitted by June 25. Stay tuned and be sure to attend the free DCI Webinar on the topic on Monday May 2 from 2:00-3:00pm EDT.

UNITED SPACE ALLIANCE CHALLENGES FINAL DOL ORDER IN FEDERAL DISTRICT COURT

by Art Gutman Ph.D., Professor, Florida Institute of Technology

In a lawsuit filed on April 19, 2011, United Space Alliance (USA) sued the DOL and OFCCP challenging a Final DOL Order for USA to produce documents or face suspension of current and future federal contracts. Yes, you read that correctly: OFCCP is being sued by a federal contractor in court (the district court for the District of Columbia).

DCI President David Cohen summarized the initial Administrative Law Judge (ALJ) ruling on this case at: http://ofccp.blogspot.com/2011/03/alj-rules-on-ofccps-desk-audit-screen.html. This is both an intriguing and important case. I will cover the major highlights.

For those that need a review: in August 2009 the OFCCP selected USA (and other companies) for compliance reviews. The OFCCP requested, among other things, USA’s affirmative action plan (AAP) and personnel activity and compensation data (hereafter called “Item 11”). The OFCCP then conducted a desk audit in which it analyzed the Item 11 data on the following three “threshold” criteria:

(1) Whether, for at least one pay division, there is a specified difference in average compensation between the groups being compared and, if so, whether at least one group appears to be adversely affected.

(2) After combining the pay divisions meeting the above condition, whether the number of employees in the nonfavored group is greater than a specified number and represents a specified percentage of the total employees in that group in the overall workforce.

(3) Whether the overall percentage of the group most adversely affected in the combined pay divisions is larger, by a specified amount, than the overall percentage of the other groups adversely affected. The specific thresholds used in each of the three criteria above

USA passed on all three criteria. Nevertheless, Miguel Rivera, Jr., District Director in the Orlando and Miami, Florida OFCCP offices did two additional “analyses”, and on that basis, requested additional data. Rivera conducted what he termed a “pattern analysis”, which included no thresholds, and a “30 and 5” test (30 or more employees in a given job with at least 5 males or 5 females) for what was termed a “refinement” analysis. Based on these analyses, Rivera then requested an additional 15 items, and ultimately three more items (for a total of 18 items). Interestingly, among the newer items were requests for AAP data relating to both Section 503 of the Rehabilitation Act and Vietnam-era Veterans’ Readjustment Assistance Act (VEVRAA).

USA refused to respond and the OFCCP brought an administrative complaint against USA that was heard by ALJ Daniel A. Sarno, Jr. (District Chief Administrative Law Judge) on February 28, 2011. During the hearing, Dr. Dr. Stephen G. Bronars, a labor economist, testified on behalf of USA to rebut the testimony of Rivera. Bronars testified that neither of Rivera’s analyses contained thresholds to eliminate minor pay differences, and that the threshold analysis was better than any analysis that did not use thresholds. However, he referred to both the threshold and non-threshold analysis as “rudimentary non-statistical” tests, and cited as the “gold standard”, regression analyses that compares “average pay, holding constant other factors that should influence pay and having the statistical methodology sort of adjust pay so that you’re not allowing factors to confound any comparisons that you would be making.” This gave Sarno the ammunition he needed to rule against USA. More specifically, he ruled:

[The] District Director was not using any rudimentary non-statistical test to charge Defendant with any violations. Rather, he was merely attempting to gather additional information to conduct a regression analysis before making any decision with regard to determining whether Defendant should be charged with any violations. I find his actions to be prudent and quite reasonable. Moreover, I find the additional 18 questions presented to Defendant to be quite limited in scope.

The rest is history. Sarno supported the OFCCP, including a ruling that there was no vindictive prosecution based on the request for Section 503 and VEVRAA data, even though he ruled that these latter requests were unreasonable. Sarno then recommended the following:

1. the Administrative Review Board (ARB) order United Space Alliance to comply with the OFCCP’s desk audit within thirty days of the ARB’s Final Decision;

2. should the ARB order United Space Alliance to comply with the on-site compliance review, the review be limited to gathering the data and/or documents related to the eighteen questions concerning possible violations of the Executive Order; and

3. should United Space Alliance fail to comply with the desk audit and/or on-site compliance review, it is recommended that the ARB cancel all of United Space Alliance’s present Federal Contracts and debar all future Federal Contracts until such time as the Federal Contractor is in compliance with the ARB’s order.

The Secretary of Labor then supported this ruling, freeing USA to sue the DOL and OFCCP in the Federal District Court of the District of Columbia.

The Secretary of Labor then supported this ruling, freeing USA to sue the DOL and OFCCP in the Federal District Court of the District of Columbia. What I find troubling is that in 2007 the OFCCP developed an internal compensation directive that spelled out the item 11 “trigger test”. Most of us referred to this as the 2-30-3 rule. The district director chose to ignore this internal protocol and instead exercised pure subjective discretion without any basis in evidence for believing there was “systemic pay discrimination”. The judge seemed to affirm OFCCP’s position that even though an internal “rule” was created for analyzing item 11, the OFCCP can essentially flip a coin when deciding on whether or not to ask for more detailed data. It will be interesting to see how this plays out in federal court.

CITY OF PHILADELPHIA PASSES ORDINANCE TO RESTRICT INQUIRIES INTO CRIMINAL HISTORY

by Art Gutman Ph.D., Professor, Florida Institute of Technology

As reported by BNA, Philadelphia PA Mayor Michael A. Nutter signed the so-called “Ban the Box” bill on April 13, 2011 prohibiting both public and private employers from divulging their criminal history until after a first employment interview. The ordinance applies to employers with 10 or more employees. Exemptions are made for criminal justice agencies (e.g., police department, courts, prisons, etc.). The stated logic behind the bill is that nearly 20% of the city’s population has a criminal record, and that by routinely conducting criminal background checks, employer set the stage for a “lifetime of discrimination” for near nearly 3,200 people that are released from the Philadelphia prison system each year. According to the bill’s statement of legislative findings and purpose, the ordinance is designed to provide job applicants with criminal records “an opportunity to be judged on his or her own merit during the submission of the application and at least until the completion of one interview”. The ordinance prohibits employers from inquiries into criminal history from the time the applicant applies to before or during the first phone or in-person interview. Employers who do not conduct interviews are prohibited from making any criminal history inquiries, and are prohibited from taking any adverse action based on arrests that do not result in conviction or pending charges. The Philadelphia City Counsel voted 13-4 to approve the ordinance, which goes into effect 90 days after the mayor signed it.

SOME THOUGHTS ON ORAL ARGUMENTS IN DUKES V. WAL MART

by Art Gutman Ph.D., Professor, Florida Institute of Technology


As reported in an alert on March 30, 2011, two of our DCI colleagues (David Morgan & Eileen Curtayne) attended oral arguments held a day earlier Wal Mart v. Dukes. I will assume the reader has read their report, which lays out the major issues in the case. In the days since, various blog writers have handicapped the oral arguments, universally agreeing that the plaintiff's arguments for certifying a class of 500,000 women on grounds of sex discrimination in pay and promotion is weak and wanting. There are basically two issues in this case, one substantive and other technical. The substantive issue is whether Wal Mart established a culture of sex-biased subjective decision-making that was common to the entire class. The technical issue, as noted by Morgan and Curtayne, is whether the 9th Circuit ruling favoring certification is more consistent with Federal Rule 23(b)(3), commonly used for monetary awards, and Federal Rule 23(b)(2), commonly used for injunctive and declaratory relief.

Often, I find the “common wisdom” weak and wanting, but not in this case. The tea leaves clearly favor Wal Mart. Basically, the three female justices (Ginsburg, Sotomayor & Kagan) were friendly to the plaintiffs on most (but not all) issues, Breyer, at best, was neutral, and predictably, Roberts, Scalia, and Alito were unfriendly, if not hostile. As usual, Thomas was quiet, but he has a history of siding with Roberts and Scalia. That leaves the one justice who generally decides potentially close rulings --- Kennedy. Kennedy lobbed a couple of softball questions at Wal Mart’s attorney (Boutrous) and bombshell questions at the plaintiff’s attorney (Sellers).

Kennedy’s unfriendliness to the plaintiffs on the substantive issue can be summed up with his first question to Sellers. Accordingly:

It's not clear to me what is the unlawful policy that Wal-Mart has adopted under your theory of the case?

To which sellers responded:

Justice Kennedy, our theory is that Wal-Mart provided to its managers unchecked discretion in the way that this Court's Watson decision addressed that was used to pay women less than men who were doing the same work in the same -- the same facilities at the same time, even though -- though those women had more seniority and higher performance, and provided fewer opportunities for promotion than women because of sex.

To which Kennedy responded:

It's -- it's hard for me to see that the -- your complaint faces in two directions. Number one, you said this is a culture where Arkansas knows, the headquarters knows, everything that's going on. Then in the next breath, you say, well, now these supervisors have too much discretion. It seems to me there's an inconsistency there, and I'm just not sure what the unlawful policy is.

Scalia then interrupted echoing the same theme:

I'm getting whipsawed here. On the one hand, you say the problem is that they were utterly subjective, and on the other hand you say there is a -- a strong corporate culture that guides all of this. Well, which is it? It's either the individual supervisors are left on their own, or else there is a strong corporate culture that tells them what to do.

Justice Roberts added:

What if you had a situation where you had a company with a very clear policy in favor of equal treatment of men and women? You know, the answer to your -- the answer to your question was women don't have as many positions because managers discriminate against them in -- in hiring and in promotion, yet you still have the same subjective delegation system. Could you have a class of women who were harmed by this subjective policy, even though it was clear that the policy of the corporation favored equal employment opportunity?

Justice Roberts also asked:

How many examples of abuse of the subjective discrimination delegation need to be shown before you can say that flows from the policy rather than from bad actors? I assume with three -- however many thousands of stores, you're going to have some bad apples.

Justice Alito, took a broader perspective, adding:

So, you have the company that is absolutely typical of the entire American workforce, and let's say every single -- there weren't any variations. Every single company had exactly the same profile. Then you would say every single company is in violation of Title VII?

I think you get the picture ….. regardless of which federal rule applies, the plaintiffs are in deep water on their substantive claim. Kennedy, Scalia, Roberts and Alito lobbed bombs at the plaintiffs, and Thomas was probably thinking the same thing.

The technical issue will also likely resolve in favor of Wal Mart. The plaintiffs favor Rule 23(b)(2) so that they can make pure statistical arguments for damages, negating any basis for Wal Mart to challenge individual claims.

Wal Mart favors Rule 23(b)(3), which is more difficult to satisfy and permits Wal Mart to challenge individual claims. There was clearly more consensus among the justices, and it favored Rule 23(b)(3). Indeed, the chief critic of Rule 23(b)(2) was Sotomayor, who stated to Sellers:

I'm -- I'm a little confused, all right? Because you're saying an individualized hearing is impossible, but that's exactly what you're saying you're going to do, only through statistics.

Sotomayor than anticipated Seller’s response and added:

You're going to say through my statistical model, I will be able to identify those women in the class who are deserving of pay raises. What that doesn't answer is when in this process is the defendant going to be given an opportunity to defend against that finding?

Stay tuned. The Supreme Court will probably rule sometime in July.