The Older-Worker-Younger-Supervisor Dyad: A Test of the Reverse Pygmalion Effect

A R T I C L E S

The Older-Worker–Younger-Supervisor Dyad: A Test of theReverse Pygmalion Effect

Mary Hair Collins, Joseph F. Hair, Jr., Tonette S. Rocco

An emerging phenomenon, the older worker reporting to a much youngersupervisor, is reversing the tradition that managers are older and moreexperienced than subordinates. These age-related demographic changes are bringing about a role reversal in the workplace that violates establishedage norms, creating status incongruence in the supervisor-subordinate dyad.This age-reversed dyad can be better understood by examining generationaldifferences and the effects of older workers’ expectations on their youngersupervisors’ leadership behavior, referred to as Reverse Pygmalion. Researchfindings highlight demographic differences and similarities in the supervisor-subordinate dyad, called Relational Demography. Major findings are thatolder workers expect less from their younger supervisors than do youngerworkers, and in turn older workers rate their younger supervisors’ leadershipbehavior lower than younger workers rate their younger supervisors.

Census numbers indicate older age groups will increase substantially in thenext few years (Fullerton & Toossi, 2001). Workforce aging is a result of the maturing of baby boomers, increased longevity, and a simultaneousdecline in the birth rate (Crampton & Hodge, 1996). Drucker (1997)believes the most important concern for businesses in the near future willnot be technology or economics, but demographics. Similarly, Reingold(1999) characterizes anticipated demographic changes as “almost likegeological plates, but it’s demographic plates. The graying of America willalter everything from office furniture to the meaning of work itself” (p. 114).

As America’s generation of baby boomers approaches retirement age,statisticians and demographers are predicting a workforce vastly different

HUMAN RESOURCE DEVELOPMENT QUARTERLY, vol. 20, no. 1, Spring 2009 © Wiley Periodicals, Inc.Published online in Wiley InterScience (www.interscience.wiley.com) • DOI: 10.1002/hrdq.20006 21

22 Collins, Hair, Rocco

HUMAN RESOURCE DEVELOPMENT QUARTERLY • DOI: 10.1002/hrdq

from any time in the past. The impact of workforce aging will be comparablein magnitude to the baby boom generation, the civil rights movement, and thewomen’s rights movement (Bronte & Pifer, 1986). Much of the currentworkforce is made up of baby boomers, followed by a considerably smallergeneration X (Venneberg, 2006). Many of the most experienced workers willsoon be eligible to retire, and there will be too few knowledgeable, skilledworkers to replace them. A gap in the supply and demand of workers isemerging—a gap that will grow to perhaps tens of millions of workers(Dychtwald, Erickson, & Morison, 2006). In addition, the workforce is becom-ing more demographically diverse, with people frequently working with otherswho differ in age, race, gender, and ethnicity (Tsui, Egan, & O’Reilly, 1992).

Changing demographics have created a shortage of skilled and expe-rienced workers, molding the workplace of the future and “shaping HR man-agement and development practices” (Stein, Rocco, & Goldenetz, 2000, p. 63).These changing demographics have increased the importance of under-standing the characteristics of older workers. For this study, older workers andolder supervisors were defined as age 50 and over. This threshold was chosenbecause it is the age of eligibility for membership in the American Associationof Retired Persons (AARP), the leading nonprofit, nonpartisan organization forolder individuals in the United States. In addition, research published byAARP (1989, 1994) defined older workers as age 50 and over.

Traditionally, managers have been older and more experienced than theirsubordinates. However, businesses are now hiring older workers for entry-level positions as well as jobs previously performed by younger workers, thusviolating traditional age norms in the workplace (Lawrence, 1988). As a resultof this demographic evolution, older workers are reporting to much youngersupervisors (Perry, Kulik, & Zhou, 1999; Shore, Cleveland, & Goldberg,2003) who were promoted into management positions because of a higherlevel of education, strategic planning expertise, or information technologyskills (Sopranos, 1999). Younger workers and supervisors were defined inthis study as age 39 and below. This age threshold is consistent with the Federal Age Discrimination in Employment Act specifying age 40 as thecutoff point between younger and older employees (Crampton & Hodge,2007). Moreover, younger workers are normally defined in the literature assomeone under 40 years of age (Smith & Harrington, 1994).

Several problems have been associated with the older-worker–younger-supervisor dyad. For example, older workers feel uncomfortable takinginstructions from supervisors the same age as their children or grandchildren(Hirsch, 1990; Shellenbarger & Hymowitz, 1994). Younger supervisors arereluctant to give orders to workers as old as their grandparents (Hirsch, 1990).Similarly, younger supervisors with older subordinates may contradict statusand age norms that suggest older, more experienced supervisors shouldsupervise younger, less experienced subordinates (Perry et al., 1999). Finally,age differences that are not compatible with workplace status norms may

The Older-Worker–Younger-Supervisor Dyad 23


negatively affect the supervisor-subordinate relationship (Tsui, Xin, & Egan,1996).

These issues have created a need for greater understanding of age-relateddemographics in the supervisor-subordinate dyad. For example, olderworkers perceive they obtain less support from younger supervisors, whileyounger supervisors believe they receive less loyalty and fewer contributionsfrom older workers (Tsui et al., 1996). These feelings may be due toincongruence in social status or violation of expected career timetable (Perryet al., 1999). Older workers also may perceive younger supervisors as havingless wisdom and experience, or as lacking the ability to obtain resources anduse upward influence in the organization (Tsui et al., 1996). Age differencesthat are not consistent with relational age norms, as with an oldersubordinate, may create negative responses such as less cooperation andlower support for younger supervisors (Tsui et al., 1996). Finally, generationaldifferences can contribute to how workers perceive the leadership of theirsupervisors (Arsenault, 2004).

This age-reversed dyad can be understood by examining the effects ofolder workers’ expectations on their younger supervisor’s leadershipbehavior—a relationship referred to as the Reverse Pygmalion effect (Eden,1984) or upward expectancy effects (Eden, 1990). The lack of research onupward expectancy effects in the workplace, as well as emerging problemsassociated with generational differences, led to this investigation of older andyounger workers’ expectations of their younger supervisors. Gilley, Eggland,and Gilley (2002) defined one mission of human resource development(HRD) as “organizational development that results in both optimal utilizationof human potential and improved human performance” (p. 13). This studyaddresses that mission by increasing HRD’s awareness of and ability torespond to this new intergenerational dyadic relationship of an older workerwith a younger supervisor, a phenomenon that is increasingly critical toimproving organizational performance.

The purpose of this research was to examine the Reverse Pygmalioneffect by focusing on upward expectancy effects in the supervisor-subordinatedyad represented by an older worker with a younger supervisor, as com-pared to a younger worker with a younger supervisor. Specifically, theupward expectancy effects examined were the subordinate’s expectations of the supervisor’s leadership behaviors. These research questions wereposed:

1. Do older workers with younger supervisors expect less effective leader-ship behaviors than do younger workers with younger supervisors?

2. Do older workers with younger supervisors expect less effective leader-ship behaviors than do younger workers with older supervisors?

3. Do older workers with younger supervisors expect less effective leader-ship behaviors than do older workers with older supervisors?



4. Do older workers with younger supervisors rate their supervisor’s leader-ship behavior lower than do younger workers with younger supervisors?

5. Do older workers with younger supervisors rate their supervisor’s leader-ship behavior lower than do younger workers with older supervisors?

6. Do older workers with younger supervisors rate their supervisor’sleadership behavior lower than do older workers with older supervisors?

Review of the Literature

As the workforce ages and younger workers are promoted into managerialpositions, generational differences will have an impact on the supervisor-subordinate relationship. Traditionally, supervisors were older and more expe-rienced, but an emerging phenomenon involves the older worker reporting toa much younger supervisor (Perry et al., 1999; Shore et al., 2003). The genera-tional differences within this new dyad can be understood better throughincreased knowledge of intergenerational dyads in the workforce. In the nextsection, we review the literature supporting the theoretical framework thatguided this research.

Supervisor-Subordinate Dyad. Early research on demographic diversityrevealed that leaders do not use the same leadership style with all subordinates,and that unique exchange relationships develop between a supervisor and each group member (Dansereau, Cashman, & Graen 1973; Graen, Novak, &Sommerkamp, 1982). These exchange relationships lead to a subordinate beingpart of either the ingroup or the outgroup, and leaders quickly categorize sub-ordinates into one of the groups (Graen & Cashman, 1975). At the same time,subordinates also categorize supervisors, contributing to the dyadic relation-ship. Tsui, Xin, and Egan (1996) studied demographic similarity and dissimi-larity in the supervisor-subordinate dyad “by drawing on a variety of theoreticalperspectives” (p. 105) and then using those theories to determine how rela-tional demography influences the quality of leader-member exchange. Theyanalyzed demographic similarities and dissimilarities and the demographiccomposition of the group, as well as ingroup and outgroup categorization.They found that both supervisor and subordinate influence the exchange rela-tionship, and in the process of categorization subordinates often associatesupervisor attributes such as age to wisdom, tenure to valuable experience, andeducation to knowledge.

Tsui and O’Reilly (1989) investigated the multivariate effects of age,education, gender, race, and tenure on supervisor’s performance ratings ofsubordinates. Their field study of 272 supervisor-subordinate dyads in aFortune 500 company supported their previous research finding thatdissimilarity within this dyad creates the perception on the part of thesupervisor of lower effectiveness and attraction to the subordinate. Theycoined the term relational demography and proposed that demographicsimilarity (or dissimilarity) between supervisor and subordinate influences

attitudes, behavior, and job performance (Tsui & O’Reilly, 1989). When thecharacteristics of individuals holding supervisory positions are inconsistentwith traditional expectations, organizational problems often emerge.

At a large Midwestern university, Perry, Kulik, and Zhou (1999) admini-stered an oral questionnaire to 35 employees during a lengthy interview. Onthe basis of hierarchical regression analysis, they found that workers whowere older than their supervisors missed less work and displayed characteri-stics of good citizenship more often than coworkers who were younger thantheir supervisors. They also found that younger supervisors contradict ageand status norms that indicate older, more experienced supervisors shouldsupervise younger, less experienced subordinates. Age differences congruentwith workplace status norms generally improve supervisor subordinaterelationships, while age differences that are incongruent can negatively affectthe relationship (Tsui et al., 1996). Status incongruence occurs when thedirection of the demographic differences causes employees to question theirworkplace hierarchy (Perry et al., 1999). For example, older workers mayperceive the younger supervisor to be lacking practical working experience,leading the older worker to be less supportive of the younger supervisor.Similarly, older subordinates experiencing status incongruence may respondnegatively to younger supervisors perceiving that their situation violates thecareer timetable typically associated with supervisory positions (Perry et al.,1999; Tsui et al, 1996). Whether attributable to age alone or cultural factorswithin age cohorts, older workers and younger supervisors can be expectedto exhibit significant dissimilarities (Collins & Fisher, 2006). Clearly, theneed to study relational demography is increasingly important as the work-force ages and traditional age norms regarding supervisor-subordinate rela-tionships are confounded.

Age Differences in the Supervisor-Subordinate Dyad. Older subordinateswith a younger supervisor often feel their supervisor does not have the knowl-edge, experience, or training to lead and mentor (Tsui et al., 1996). Even when younger supervisors have advanced education, older subordinatesbelieve they lack experience and as a result support them less. At the sametime, younger supervisors believe older workers lack management knowledgefrom advanced studies. Demographic attributes can have a negative effect onthe supervisor-subordinate exchange relationship, especially attributes that areimmediately visible such as gender, race, age, and other physical qualities (Tsuiet al., 1996). Older supervisors typically are viewed as having more experienceand wisdom, while younger, less experienced supervisors disrupt the exchangerelationship because older subordinates often view them less favorably. Thisrole reversal violates established age norms that older workers are more expe-rienced and supervise younger, inexperienced workers (Perry et al., 1999).

Employee perceptions of demographic differences such as age areimportant because “employees’ beliefs, whether or not they are consistentwith reality, affect their behaviors” (Ensher, Grant-Vallone, & Donaldson,




2001). How a person perceives the world influences the way he or she reacts toevents (Calhoun, Cann, Tedeschi, & McMillan, 1998), and a perception that aworker is either similar or dissimilar in demographic characteristics to othermembers of the workgroup influences work-related outcomes (Chattopadhyay,Glick, Miller, & Huber, 1999). The perceptions older workers have of theirsupervisors affect their expectations of the younger supervisor’s leadershipcapability. In fact, some researchers and practitioners believe that perceptionis reality, and when looking at survey research it is based solely on percep-tions (Waclawski, 2002).

Generational Differences. Because of differing attitudes, preferences, andvalues, generational differences also create incongruence in the supervisor-subordinate dyad. Generational groups experience historical events that impartpowerful memories, molding beliefs about institutions and authority (Conger,2001). Beliefs in turn influence workplace attitudes and contribute to howeach generation perceives the leadership of its supervisors (Arsenault, 2004).For example, differences exist between baby boomers and generation Xers intheir commitment to organizations (Daboval, 1998) as well as in the impor-tance they place on trust (Robinson & Jackson, 2001). With regard to leader-ship, baby boomers prefer competent, caring, and honest leaders, while Xersprefer leaders who challenge the system and create change (Arsenault, 2004).In addition, baby boomers consider honesty very important and want leadersto be open about their values and ethics, while honesty and ethics are lessimportant to Xers. Finally, boomers are known for their strong work ethic andloyalty to the organization. But Xers see their parents as sacrificing all to theircareers just to become victims of corporate downsizing, causing Xers to be cyn-ical, untrusting, and less loyal (Davis, Pawlowski, & Houston, 2006). Becauseof these generational differences, some older workers stick together, becomingmuch closer than younger workers (Mott, 2006).

Downward and Upward Expectancy Effects in Dyads. Expectations ofboth the supervisor (downward) and the subordinate (upward) also influencethe supervisor-subordinate relationship. The term “Pygmalion effect” wascoined to represent a person acting on the expectations of another (Rosenthal &Jacobsen, 1968). The Pygmalion effect proposes that higher expectations in aleader-follower relationship elicit more effective performance, while lowerexpectations elicit less effective performance. Similarly, expectancy theoryposits that people choose to perform activities they believe they have a highexpectation of performing successfully (Vroom, 1964). Within an organiza-tional context, the Pygmalion effect occurs when supervisory expectations have an impact on the performance of subordinates. At the same time, theReverse Pygmalion effect occurs when subordinate expectations affect the per-formance of supervisors (Eden, 1984).

The initial study of the Pygmalion effect demonstrated that pupils exhibi-ted more progress when their teachers expected more of them (Rosenthal &Jacobsen, 1968). In their classic, empirical study at Oak School, they confirmed


that pupils from whom teachers expected greater intellectual gain actuallyadvanced more than the control group. In a later theoretical analysis of previousPygmalion research, Murphy, Campbell, and Garavan (1999) found that whenteachers believed children were underperformers, they no longer spent timeteaching those children, and the children later performed poorly in academicstudies. Studies have also been conducted to examine the Pygmalion effect inthe supervisor-subordinate relationship in organizations. In a field experimentof 105 soldiers in the Israeli Defense Forces, Eden and Shani (1982) used ablock experimental design to induce different levels of expectancy regardingtrainees’ potential to command. They measured “trainees’ learning performance,attitudes and their perceptions of instructor leadership” (p. 196). By analyzingthe means for each level of instructor expectancy, they found that instructorexpectations influenced trainee performance, thus confirming that leader behaviorin organizations is similar to that of teachers. In a meta-analysis of Pygmalionexperiments, Eden (1984) reviewed previous studies on the Pygmalion effect inmultiple contexts. After analyzing previous studies, he concluded that thePygmalion effect could be a useful management tool in raising productivity(Eden, 1984, 1990, 2003).

A Reverse Pygmalion effect exists when subordinate performance andexpectations influence the leadership behavior of supervisors. In anempirical investigation in a workplace setting, Lowin and Craig (1968)found that the performance of subordinates influenced the leadership oftheir superiors. They deceptively hired subjects for supervisory positions inan office who believed they had actually been hired, but they were part of afield study where they were observed as they reacted to predeterminedsituations set up with either competent or incompetent subordinates. Theyconcluded from their study that the performance of subordinates shapes theleadership styles of their supervisors in how closely the supervisors monitorthe subordinates, how much the supervisor emphasizes organizational goals,and consideration for subordinates. Finally, Eden (1984) found that asubordinate’s high performance could arouse high expectations in thesupervisor, resulting in better supervisor leadership.

The Reverse Pygmalion theory, later referred to as upward expectancyeffects (Eden, 1990), further proposes that highly performing subordinatesimprove supervisory leadership, leading to performance improvements inthe organization. Similarly, subordinates’ high expectations of their supervisorevoke more effectiveness from the supervisor, just as lower expectationselicit less effective performance. Thus subordinates can be taught to behavein a manner that evokes more effective leadership from supervisors, therebytriggering Pygmalion in reverse. Research on the Pygmalion effect has mainlybeen top-down, involving the expectations of a person in authority about theperformance of lower-level subordinates. In contrast, expectations ofsubordinates toward their supervisor, referred to as bottom-up, have beenlargely ignored (Eden, 1990).




Research Method

This section first describes the study measures and data collection approach.It then addresses preliminary analysis of the data to facilitate examining theresearch questions.

Study Measures. A questionnaire was developed using constructs from thesupervision literature, input from knowledgeable experts in the field, and inter-views with managers and employees. The two leadership constructs used wereworker expectations of their supervisors’ leadership and worker perceptionsof the supervisor’s leadership behavior. Leadership expectations were measuredby the Leadership Expectations Inventory (LEI) developed by Gurie (2002).With the original LEI instrument, respondents rated their supervisor on 12items using a five-point Likert scale from Never to Always. Examples of items:I expect my supervisor to be “overall a strong leader,” I expect my supervisorto be “an effective communicator,” and I expect my supervisor to be “a goodencourager.” The Cronbach alpha for the original LEI instrument was .96. Theoriginal LEI was adapted for this study because two questions were double-barreled. For the present study, these double-barreled items were separatedinto two separate items. This was done because it is impossible to know whichof the two adjectives a respondent is reacting to, and respondents would notknow how to answer if they have differing opinions about the two descriptors(Hair, Money, Samouel, & Page, 2007). In addition, the adjective helpful wasadded to the construct as another item because this dimension was suggestedin preliminary in-depth interviews as a component of the supervisory expec-tations domain. Separation of the two double-barreled items and the additionof the “helpful” item brought the total number of LEI items to 15 in this study.Reliability of the revised expectations measure was assessed with Cronbach’salpha, which revealed an a� .97. The sample used to assess the reliability wasthe 319 older and younger workers and supervisors on which the study’s find-ings are based.

The face validity of the LEI was established on the basis of input from in-depth interviews of managers and workers, as well as psychometric experts.Exploratory factor analysis (EFA) was used to further assess the validity of theLEI. EFA was chosen because it yields a direct picture of dimensionality(Hurley, Scandura, Schreisheim, Brannick, Vandenberg, & Williams, 1997).The EFA approach was common factor analysis because it is more appropriatethan principal components analysis when the objective is to identify latentstructures (Nunnally & Bernstein, 1994). Moreover, the Kaiser’s measure ofsampling adequacy was .95, which means the data were appropriate for anexploratory common factor analysis. An oblique rotation was used because itis also more appropriate for latent variable investigation when latent variablesare expected to have some correlation (Hair, Black, Babin, Anderson, &Tatham, 2006). The sample used for the EFA was the 319 older and youngerworkers and supervisors on which the study’s findings are based, which


represented a 21.2:1 respondent-to-item ratio. Generally a ratio of between5:1 and 10:1 is desirable (Hair et. al., 2006), so this sample would be consid-ered to have strong respondent-to-item ratios.

The Leadership Practices Inventory-Observer version (LPI-O) was usedto assess workers’ perceptions of their supervisor’s leadership behavior(Kouzes & Posner, 2003). The LPI-O has 30 items measuring five leadershipbehavior attributes identified using exploratory factor analysis: Modeling theWay, Inspiring a Shared Vision, Challenging the Process, Enabling Others toAct, and Encouraging the Heart. This instrument was developed from morethan 1,000 case studies, followed up by 38 in-depth interviews to determinewhat leaders did in their personal best experiences as leaders (Posner &Kouzes, 1990). Results from the initial qualitative findings were thenanalyzed quantitatively with another sample of more than 2,100 managersand their subordinates (Posner & Kouzes, 1988). The LPI-O has excellentface validity, and construct validity was further established by a study thatconcluded the five leadership behavior attributes were significantly related tosubordinates’ rating of managerial effectiveness; internal reliability rangedfrom .81 to .92, and test-retest reliability ranged from .93 to .95 (Posner &Kouzes, 1992). Reliability of the LPI-O for the current study was assessedusing the 319 older and younger workers and supervisors. The Cronbach’salpha measure of reliability was .98. For purposes of consistency in datacollection across the two constructs in this study, a 10-point Likert scale wasused on both the LEI and LPI-O, with 10 being “Strongly Agree” and 1 being“Do Not Agree at All.” A 10-point Likert scale was used because it increasesthe precision in the scaled responses and reduces bias when respondentstend to avoid extreme scale points (Hair et. al., 2007).

The final questionnaire included a total of 45 items for the LEI and LPI-O.In addition, demographic information such as respondent age, gender,educational level, and firm size were obtained. The questionnaire was pre-tested on a small sample (N � 11) of subordinates and supervisors to examineclarity of instructions and sequence and understanding of questions.

Data Collection and Preparation. The questionnaire was uploaded to aWebsite by a professional research firm. The firm maintains a list of individu-als employed by a cross-section of U.S. businesses that have agreed to partic-ipate in surveys but are not compensated. The firm solicited participation bysending a notice to the list of individuals inquiring if they would be interestedin responding to a survey focusing on the relationships between workers andtheir immediate supervisor. The individuals were informed that the surveywould take approximately 10 minutes to complete and that their responseswould be totally anonymous. A total of 1,500 individuals were invited to takethe survey. The final response was 566, for a response rate of 37.7 percent.

The responses were coded to create four analysis groups, as shown inTable 1. The four groups were older worker, younger supervisor; older worker,older supervisor; younger worker, younger supervisor; and younger worker,




older supervisor. For this study, older workers and older supervisors weredefined as age 50 and above (AARP, 1989, 1994) and younger workers andsupervisors as age 39 and below (Smith & Harrington, 1994). These agedifferences ensured that comparisons were based on demographic differencesand consistent with research on relational demography that investigates demo-graphic differences, such as age, between an employee and another member ofhis or her workgroup (Perry et al., 1999). Because “differences in the attitudes,values, and beliefs of each generation affect how each generation viewsleadership” (Arsenault, 2004), the stated age differences ensured data werecollected from distinct generations.

Note that to create the analysis groups responses identifying workers orsupervisors aged 40 to 49 were removed from the sample. Thus no middle-agedyadic relationships such as older-worker–middle-age supervisor, or middle-age-worker–younger supervisor were examined. This ensured that in all com-parisons the age difference between groups was a minimum of 10 years. Thejustification for removing this middle group was in-depth interviews withworkers and supervisors in a preliminary study. These individuals indicatedthat perceptions of an older or younger relationship with another individualwere not manifested unless several years’ difference existed in ages. Theseexpressed perceptions along with the study’s definition of older and youngerworkers/supervisors led to this group being removed from the analysis.

Results reported in this study are based on comparisons of the sub-groups of the 319 individuals in the worker-supervisor groups aged 39 oryounger, and 50 or older. The sizes of the older-worker–younger-supervisorand older-worker–older-supervisor groups are somewhat smaller than theother two groups. The lower representation of these groups in the sample is a reflection of the proportion they represent in the worker-supervisorpopulation as a whole. That is, these two groups have emerged only inrecent years and still represent a relatively smaller proportion of the overall


Table 1. Respondent Groups for Analysis

Analysis Group Age Characteristics N

Older Worker – Younger Supervisor Older Worker � 50� with 45(OW-YS) Younger Supervisor � 39 or less

Older Worker – Older Supervisor Older Worker � 50� with 51(OW-OS) Older Supervisor � 50�

Younger Worker – Younger Supervisor Younger Worker � 39 or less with 143(YW-YS) Younger Supervisor � 39 or less

Younger Worker – Older Supervisor Younger Worker � 39 or less with 80(YW-OS) Older Supervisor � 50�

Total 319

worker-supervisor population. Although they are somewhat smaller, theywere considered sufficiently large to represent these two groups. Table 1presents the respondent groups, their ages, and sample sizes.

A demographic profile of the responses included in the analysis isshown in Table 2. The average age of the younger worker groups was 27 andthe average age of the older worker groups was 54. Approximately 70% ofrespondents were female and 30% male; about 80% were employed in awhite-collar position. Respondents were asked to evaluate their immediatesupervisor, so demographic characteristics for the immediate supervisor werealso obtained. The average age of younger supervisors was 34 and the averageage for older supervisors was 55. The typical educational level for both respon-dents and immediate supervisors was an undergraduate degree, althoughabout 20% of the immediate supervisors had a graduate degree. The averagesize of the respondents’ firms was about 8,500 employees.

Limitations. One limitation of the study is that the data were obtainedfrom a judgment sample instead of randomly. Because an established researchfirm collected data via the Internet, the most frequently used data collectionmethod in the United States (Research Industry Trends, 2006), we must con-sider the ability to generalize the findings. Respondents reported working forboth large and small companies, were located in more than half of the 50states, and met the criteria to define the older-worker–younger-supervisordyad. But one must still be cautious in inferring the results of this sample tothe population. Another limitation is the fact that supervisor-subordinate dyadswere not matched, and qualitative data were not collected about perceptionsof relationships between members of dyads.

Data Analysis and Results

Table 3 contains the means and correlations of the study variables. Overallsummated scores are reported for the LEI and LPI-O. Individual summated



Table 2. Demographic Characteristics of Study Respondents

Respondenta Immediate Supervisor

Groups Age Typical Education Level Age Typical Education Level

1. YW-YS 26 Undergraduate degree 33 Undergraduate degree

2. YW-OS 28 Undergraduate degree 55 Undergraduate degree

3. OW-YS 55 Undergraduate degree 35 Undergraduate degree

4. OW-OS 53 Undergraduate degree 55 Undergraduate degree

Firm size: the average size of firm that respondents worked for was about 8,500 employees.

Respondent gender: about 70% of respondents were female and 30% male.

Type of job: about 80% of the respondents were employed in a white-collar position.aRespondents were workers and supervisors, either younger age � 39 years or older � 50 years.


scores for the five composite leadership attributes making up the LPI-O alsoare shown. As can be seen, the mean level of leadership expectations of theimmediate supervisor for the total sample is quite high: 8.23 on a 10-pointscale, with 10 being the highest level of expectations. The overall evaluationof the immediate supervisor’s leadership practices is somewhat above themidpoint of the 10-point scale (mean � 6.13). Two of the compositeleadership attributes (Inspires Shared Vision and Challenges the Process)were rated below the mean of 6.13 and three (Models the Way, EnablesOthers to Act, and Encourages the Heart) were rated above the mean.

All of the relationships are positively and significantly correlated. Thusworker expectations of their immediate supervisors’ leadership effectivenessare associated with their perceptions of their leadership practices. That is, if subordinates expect a higher level of performance from their immediatesupervisor they also perceive that their supervisor performs relatively higher.Similarly, if subordinates’ performance expectations are lower for theirimmediate supervisor then perceptions of leadership performance are lower.

To examine the research questions, data analysis required testing forstatistical differences between the means of two metric variables from twosample groups, workers versus supervisors. For all research questions, thenull hypothesis was no differences between the two groups being tested.These considerations indicated that analysis of variance (ANOVA) was theappropriate statistical technique to examine the research questions (Hair et al.,2007).

Research Question 1: Do older workers with younger supervisors expect lesseffective leadership behaviors than do younger workers with youngersupervisors?Table 4 presents the ANOVA tests comparing the leadership expectations

of two groups. The mean expectations level of older workers withyounger supervisors was 6.97, whereas the mean expectations level of


Table 3. Descriptive Statistics and Correlations Among Study Variables (N � 319)

Variables Mean 1 2 3 4 5 6

1. Expectations (LEI) 8.232. Management Behaviors (LPI-O) 6.13 .436*3. Models the Way 6.15 .410* .939*4. Inspires Shared Vision 5.84 .376* .920* .833*5. Challenges the Process 5.83 .399* .948* .852* .894*6. Enables Others to Act 6.74 .452* .926* .853* .767* .832*7. Encourages the Heart 6.12 .401* .943* .861* .808* .858* .876*

Note: *p � 0.001.

younger workers with younger supervisors was 8.31. The differencebetween the means of the two groups, 1.34, was statistically significantat the .000 level. Thus older workers with younger supervisors expectless effective leadership behaviors than do younger workers withyounger supervisors.

Research Question 2: Do older workers with younger supervisors expect less effective leadership behaviors than younger workers with oldersupervisors?Table 4 presents the ANOVA tests comparing leadership expectations of

two groups. The mean expectations level of older workers with youngersupervisors was 6.97, whereas the mean expectations level of youngerworkers with older supervisors was 8.61. The difference between thetwo groups, 1.64, was statistically significant at the .000 level. Weconclude, therefore, that older workers with younger supervisors doexpect less effective leadership behaviors than younger workers witholder supervisors.

Research Question 3: Do older workers with younger supervisors expect lesseffective leadership behaviors than older workers with older supervisors?Table 4 presents the ANOVA tests comparing leadership expectations of

two groups. The mean expectations level of older workers with youngersupervisors was 6.97, whereas the mean expectations level of youn-ger workers with younger supervisors was 8.49. The difference betweenthe two groups, 1.52, was statistically significant at the .000 level. Weconclude, therefore, that older workers with younger supervisors do expect less effective leadership behaviors than older workers with oldersupervisors.



Table 4. ANOVA for Expectations of Immediate Supervisors’ Leadership Effectiveness

Leadership Expectations (LEI)

Standard MeanGroups N Means Deviations Difference Sig. Effecta

Research OW-YS 45 6.97 2.568 1.34 .000 .624Question 1 YW-YS 143 8.31 1.620Research OW-YS 45 6.97 2.568 1.64 .000 .830Question 2 YW-OS 80 8.61 1.100Research OW-YS 45 6.97 2.568 1.52 .000 .742Question 3 OW-OS 51 8.49 1.340

aCohen’s d � .5 is medium effect and .8 is large effect.


Research Question 4: Do older workers with younger supervisors rate theirsupervisor’s leadership behavior lower than do younger workers withyounger supervisors? Table 5 presents the ANOVA tests comparing evaluations of immediate

supervisors’ leadership practices for the two groups. The meanperformance level of older workers with younger supervisors was 4.67,whereas the mean performance level of younger workers with youngersupervisors was 6.41. The difference between the means of the twogroups, 1.74, was statistically significant at the .000 level. Thus olderworkers evaluate their younger supervisors’ leadership behavior lowerthan do younger workers with younger supervisors.

Research Question 5: Do older workers with younger supervisors rate theirsupervisor’s leadership behavior lower than do younger workers witholder supervisors?Table 5 presents the ANOVA tests comparing evaluations of immediate

supervisors’ leadership practices for the two groups. The meanperformance level of older workers with younger supervisors was 4.67,whereas the mean performance level of younger workers with oldersupervisors was 6.43. The difference between the two groups, 1.76, wasstatistically significant at the .000 level. Thus older workers withyounger supervisors rate their supervisors’ leadership behavior lowerthan younger workers with older supervisors.

Research Question 6: Do older workers evaluate their younger supervisor’sleadership behavior lower than do older workers with older supervisors?


Table 5. ANOVA for Evaluations of Immediate Supervisors’ Leadership Performance

Leadership Performance (LPI-O)

Standard MeanGroups N Means Deviations Difference Sig. Effecta

Research OW-YS 45 4.67 2.418 1.74 .000 .732Question 4 YW-YS 143 6.41 2.331Research OW-YS 45 4.67 2.418 1.76 .000 .778Question 5 YW-OS 80 6.43 2.095Research OW-YS 45 4.67 2.418 1.53 .000 .663Question 6 OW-OS 51 6.20 2.193

aCohen’s d � .5 is medium effect and .8 is large effect.

Table 5 presents the ANOVA tests comparing evaluations of immediatesupervisors’ leadership practices for the two groups. The meanperformance level of older workers with younger supervisors was 4.67,whereas the mean performance level of older workers with oldersupervisors was 6.20. The difference between the two groups, 1.53, wasstatistically significant at the .000 level. Thus older workers evaluatetheir younger supervisors’ leadership behavior lower than older workerswith older supervisors.

Conclusions and Discussion

The overall workforce is growing older, but the proportion of youngersupervisors is increasing. Although negative stereotypes have been associatedwith older workers, research supports their positive attributes includingdependability, loyalty, high work ethic, exemplary attendance, and goodcitizenship (Perry et al., 1999; Rix, 1997; Ramsey, 2003). Workplace changesshow an increase in older workers reporting to younger supervisors, thuscreating a need for research on supervisor-subordinate relationships. Inaddition, studies of Pygmalion and Reverse Pygmalion effects have shownthat supervisory expectations have an impact on the performance of thesubordinate (Eden, 1984) and that subordinates’ expectations of theirsupervisor can affect the supervisor’s leadership behavior (Eden, 1990).

The Pygmalion effect proposes that higher expectations in a supervisor-subordinate relationship elicit more effective performance and lowerexpectations elicit less effective performance (Eden, 1984). Within thecontext of this research study, the Pygmalion effect occurs when supervisoryexpectations influence the performance of subordinates. Similarly, the ReversePygmalion effect occurs when subordinate expectations have an impact on the performance of supervisors. The major findings of this study are that older workers expect less from their younger supervisors than doyounger workers, and in turn older workers rate their younger supervisors’leadership behavior lower than younger workers rate their younger super-visors, thus confirming the Reverse Pygmalion effect. In addition, olderworkers expect less from their younger supervisors than do younger workerswith older supervisors or older workers with older supervisors, and theyalso rate their younger supervisors’ leadership behavior lower than youngerworkers with older supervisors or older workers with older supervisors.Again, this confirms the Reverse Pygmalion effect because employee expec-tations affect how they rate the leadership behavior of their supervisors.

Summary and Future Research. The major findings of this study are thatolder workers expect less from their younger supervisors than do youngerworkers, and in turn older workers rate their younger supervisors’ leadershipbehavior lower than younger workers rate their younger supervisors. These




findings are consistent with Byrne’s research (1971) showing that individualsview people more favorably who are similar in attitudes and experience. Gen-erational studies indicate older workers have a high work ethic, take pride intheir work achievements, and think retirement would be boring (Ramsey,2003). In contrast, younger workers distrust big business, are less loyal, andleave the office promptly at 5:00 every day (Gardner, 2005). The results of ourstudy are likely related to these generational differences, reflecting older work-ers’ disillusionment with the contrasting work ethic of the younger generation.For example, older workers evaluated their younger supervisors’ performanceless favorably. This finding is consistent with previous studies showing theexchange relationship in the supervisor-subordinate dyad is negatively affectedby demographic differences, especially attributes that are immediately visiblesuch as age (Tsui et al., 1996). If the lower expectations of older workers con-tributed to their younger supervisors’ lower performance, our research showsthat expectations can influence outcomes and confirms the Reverse Pygmalioneffect in the older-worker–younger-supervisor dyad.

This study represents one of the first attempts to examine the ReversePygmalion effect using individuals currently experiencing this phenomenon. Asan initial effort in this emerging area, our results offer new insights as well as abasis for replication and further refinement. Future research should addressthese generational differences and identify the circumstances surrounding thenegative experiences, thus leading to increased understanding of the potentialperformance effects of expectancies on the supervisor-subordinate dyad. Betterunderstanding of the potential performance effects of expectancies on thesupervisor-subordinate dyad should lead to further research on the older workerwith a younger supervisor. Moreover, as the workforce continues to grow older,further research into all aspects of the older worker will lead to greaterutilization of this important segment of the workforce.

Implications for HRD Research and Practice. HRD has become a dynamicforce in bringing about change to greatly enhance organizational performance.Gilley, Eggland, and Gilley (2002) defined the mission of HRD as offering indi-vidual development, career development, performance management, and orga-nizational development. These four characteristics of HRD’s mission areimportant to HRD professionals in all organizations. Therefore they are dis-cussed individually in relation to our research findings.

“Individual development [is] focused on performance improvementrelated to a current job” (Gilley, Eggland, & Gilley 2002, p. 12). In consider-ing the impact of individual development, HRD professionals have expressedconcern about an aging workforce and its influence on organizational perfor-mance. From a research perspective, this study confirms the importance offocusing on the individual development of both the older worker and theyounger supervisor. Individual knowledge of the power of expectations in the supervisor-subordinate dyad could contribute to a better workforce andto overall improvements in the organization. Our findings also suggest HRD


professionals should develop training programs for younger supervisors todevelop and improve their supervisory skills in managing an older popula-tion. Because perceptions may influence the subordinate’s belief about theleadership ability of the supervisor (Tsui et al., 1996), and “such perceptionsrepresent employees’ thinking about such issues as relationships with theirsupervisors” (McMurray, Scott, & Pace, 2004), HRD professionals and practi-tioners need to develop training programs to address issues surrounding the qualifications of younger supervisors by focusing on development ofboth older workers and younger supervisors. These programs should assistolder workers in better understanding younger supervisors, thus improvingtheir negative perceptions. They should specifically focus on training bothyounger supervisors and older workers about cultural and generationaldifferences. This training will lead to greater understanding within this dyad.As HRD professionals focus on developing and training the individual, bothyounger supervisors and older workers will gain the tools needed to focuson the development of their respective careers.

“Career development [is] related to future job assignments” (Gilley et al.,2002, p. 12). This study also confirms the value of career developmentrelated to future job assignments. HRD professionals familiar with upwardexpectancy effects should leverage this knowledge to increase performanceby designing and implementing career development training on the power ofexpectations to bring about more effective leadership. Career developmenttraining should include information on age differences and the value eachgeneration contributes to the workplace, thus increasing the older worker’sexpectations of their younger supervisor and leading to higher performanceon the part of younger supervisors. If employees are given the opportunityto develop and advance their careers, this will in turn lead to better system-wide performance.

“Performance management systems [are] used to enhance organizationalperformance capacity and capability” (Gilley et al., 2002, pp. 12–13).According to Gilley et al., performance management is an approach thatimproves organizational performance by focusing on the total organizationalsystem. Organizations commonly use leadership development programs toimprove systemwide organizational performance from the top down. Byusing the findings of our research, HRD professionals should design leader-ship development programs that emphasize generational differences bymaking training programs more relevant to the younger generational cohorts(Arsenault, 2004). They should also focus on aspects of management andleadership that can develop younger supervisors. In addition, systemwideperformance improvements can be attained by planning for and systema-tically employing an age-diverse workforce (Jayne & Dipboye, 2004).Competitive readiness and profitability of the whole organization as a systemcan be enhanced as HRD professionals find ways to use the diverse talentcreated by this emerging dyad of an older worker with a younger supervisor.




This readiness and profitability can be further enhanced if HRD professionalsimplement plans for organizational development.

“Organizational development that results in both optimal utilization ofhuman potential and improved human performance . . . [is] measured byincreased competitive readiness, profitability, and renewal capacity” (Gilley et al., 2002, p. 13). Our study confirms the importance of this fourth compo-nent of HRD’s mission. Greater understanding of this new intergenerationaldyad will increase our ability to respond to a dyadic relationship critical toimproving organizational development and performance. Moreover,organizations can enhance the value of both younger supervisors and oldersubordinates by designing relevant training to enhance the development ofeach member of the supervisor-subordinate dyad. An understanding of thepower of expectations in the supervisor-subordinate dyad gives HRD profes-sionals a new performance lever to use in designing innovative trainingmethods, thus developing a company’s best resource: its employees.

References

American Association of Retired Persons. (1989). Business and older workers: Current perceptionsand new directions for the 1990s. Washington, DC: Andrus Foundation.

American Association of Retired Persons. (1994). Valuing older workers: A study of costs and pro-ductivity. Washington, DC: ICF Corporation.

Arsenault, P. M. (2004). Validating generational differences: A legitimate diversity and leadershipissue. Leadership & Organization Development Journal, 25(1/2), 124–141.

Bronte, L., & Pifer, A. (Eds.). (1986). Our aging society: Paradox and promise. New York: CarnegieCorporation.

Byrne, D. (1971). The attraction paradigm. New York: Academic Press.Calhoun, L. G., Cann, A., Tedeschi, R. G., & McMillan, J. (1998). Traumatic events and genera-

tional differences in assumptions about a just world. Journal of Social Psychology, 138(6),789–791.

Chattopadhyay, P., Glick, W., Miller, C. C., & Huber, G. (1999). Determinants of executive beliefs:Comparing functional conditioning and social influence. Strategic Management Journal, 20,763–789.

Collins, M. H., & Fisher, S. R. (2006). Harnessing Pygmalion in reverse. In F. M. Nafukho (Ed.),Proceedings of the Academy of Human Resource Development 2006 Annual Conference(pp. 1332–1339). Columbus, OH: AHRD.

Conger, J. (2001). How ‘Gen X’ manages. In J. Osland, D. Kolb, & I. W. Rubin (Eds.), Organiza-tional Behavior Reader (pp. 9–19). Upper Saddle River, NJ: Prentice Hall.

Crampton, S., & Hodge, J. (1996). Transition—ready or not: The ageing of America’s work force.Public Personnel Management, 25(2), 243–256.

Crampton, S. M., & Hodge, J. W. (2007). Age discrimination and downsizing. Business Review,Cambridge, 7(1), 341–347.

Daboval, J. M. (1998). A comparison between Baby Boomer and Generation X employees’ bases andfoci of commitment. Doctoral dissertation, Nova Southeastern University.

Dansereau, F., Jr., Cashman, J., & Graen, G. (1973). Instrumentality theory and equity theory ascomplementary approaches in predicting the relationship of leadership and turnover amongmanagers. Organizational Behavior and Human Performance, 10, 184–200.


Davis, J. B., Pawlowski, S. D., & Houston, A. (2006). Work commitments of Baby Boomers andGen-Xers in the IT profession: Generational differences or myth? Journal of Computer Informa-tion Systems, 46(3), 43–49.

Drucker, P. E. (1997). The future that has already happened. Harvard Business Review, 75(5),20–22.

Dychtwald, K., Erickson, T. J., & Morison, R. (2006). Workforce crisis. Boston: Harvard BusinessSchool Press.

Eden, D. (1984). Self-fulfilling prophecy as a management tool: Harnessing Pygmalion. Academyof Management Review, 9(1), 64–73.

Eden, D. (1990). Pygmalion in management. Lexington, MA: Lexington Books.Eden, D. (2003). Self-fulfilling prophecies in organizations. In J. Greenburg (Ed.), Organizational

behavior: The state of the science (2nd ed., pp. 91–124). Mahwah, NJ: Erlbaum.Eden, D., & Shani, A. B. (1982). Pygmalion goes to boot camp: Expectancy, leadership, and

trainee performance. Journal of Applied Psychology, 67(2), 194–199.Ensher, E. A., Grant-Vallone, E. J., & Donaldson, S. I. (2001). Effects of perceived discrimina-

tion on job satisfaction, organizational commitment, organizational citizenship behavior, andgrievances. Human Resource Development Quarterly, 12(1), 53–72.

Fullerton, H. N., Jr., & Toossi, M. (2001). Labor force projections to 2010: Steady growth andchanging composition. Monthly Labor Review, November, 21–38.

Gardner, M. (2005, January 3). Meet the young(er) boss: Older Americans are increasingly likelyto work for someone younger. Christian Science Monitor, p. 13.

Gilley, J. W., Eggland, S. A.. & Gilley, A. S. (2002). Principles of human resource development. Cam-bridge, MA: Perseus.

Graen, G., & Cashman, J. (1975). A role-making model of leadership in formal organizations: Adevelopment approach. In J. G. Hunt & L. L. Larson (Eds.), Leadership frontiers (pp. 143–165).Kent, OH: Kent State University Press.

Graen, G., Novak, M. A., & Sommerkamp, P. (1982). The effects of LMX and job design on pro-ductivity and satisfaction: Testing a dual attachment model. Organizational Behavior and HumanPerformance, 30, 109–131.

Gurie, J. R. (2002). The relationship between perceived leader behavior and alcohol consumption amonguniversity students who are members of social fraternities. Doctoral dissertation, Louisiana StateUniversity, (URN etd-0417102-2 01343).

Hair, J. F., Jr., Black, W. C., Babin, B., Anderson, R., & Tatham, R. (2006). Multivariate data analy-sis. Upper Saddle River, NJ: Prentice Hall.

Hair, J. F., Jr., Money, A. H., Samouel, P., & Page, M. (2007). Research methods for business.Hoboken, NJ: Wiley.

Hirsch, J. S. (1990, February 26). Older workers chafe under young managers. Wall Street Jour-nal, p. B1.

Hurley, A. E., Scandura, T. A., Schreisheim, S. A., Brannick, A. S., Vandenberg, R. J. & Williams,L. J. (1997). Exploratory and confirmatory factor analysis: Guidelines, issues and alternatives.Journal of Organizational Behavior, 18, 667–683.

Jayne, M.E.A., & Dipboye, R. L. (2004). Leveraging diversity to improve business performance:Research findings and recommendations for organizations. Human Resource Management, 43(4),409–424.

Kouzes, J. M., & Posner, B. Z. (2003). Leadership practices inventory facilitator’s guide (3rd ed.).San Francisco: Pfeiffer.

Lawrence, B. S. (1988). New wrinkles in the theory of age: Demography, norms, and performanceratings. Academy of Management Journal, 31(2), 309–337.

Lowin, A., & Craig, J. R. (1968). The influence of level of performance on managerial style: Anexperimental object-lesson in the ambiguity of correlational data. Organizational Behavior andHuman Performance, 3, 440–458.




McMurray, A. J., Scott, D. R., & Pace, R. W. (2004). The relationship between organizational com-mitment and organizational climate in manufacturing. Human Resource Development Quarterly,15(4), 473–488.

Mott, V. W. (2006). Working late: Exploring the workplace, motivations, and barriers of work-ing seniors. In T. S. Rocco & J.G.L. Thijssen (Eds.), Older workers, new directions: Employmentand development in an ageing labor market (pp. 158–167). Miami: Center for Labor Researchand Studies, Florida International University.

Murphy, D., Campbell, C., & Garavan, T. N. (1999). The Pygmalion effect reconsidered: Its impli-cations for education, training and workplace learning. Journal of European Industrial Training,23(4/5), 238–250.

Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory. New York: McGraw-Hill.Perry, E. L., Kulik, C. T., & Zhou, J. (1999). A closer look at the effects of subordinate-supervisor

age differences. Journal of Organizational Behavior, 20, 341–357.Posner, B. Z., & Kouzes, J. M. (1988, Summer). Development and validation of the leadership

practices inventory. Educational and Psychological Measurement, 48, 483–496.Posner, B. Z., & Kouzes, J. M. (1990). Leadership practices: An alternative to the psychological

perspective. In K. E. Clark & M. B. Clark (Eds.), Measures of leadership. West Orange, NJ: Lead-ership Library of America.

Posner, B. Z., & Kouzes, J. M. (1992). Person-organization values congruence: No support forindividual differences as a moderating influence. Human Relations, 45(4), 351–361.

Ramsey, R. D. (2003, June). Tapping the strengths of older workers. SuperVision, 64(6), 9–12.Reingold, J. (1999, September 20). Brain drain. BusinessWeek, 113–126.Research industry trends 2006 report. (2006). Pioneer Marketing Research, Summit, NJ: Green-

book, Dialtek & Rockhopper Research.Rix, S. E. (1997). IMR 101: Workforce age composition. Comprehensive issue management report.

St. Petersburg, FL: Human Resource Institute, Eckerd College.Robinson, R. V., & Jackson, E. F. (2001). Is trust in others declining in America? An age-period-

cohort analysis. Social Science Research, 30, 117–145.Rosenthal, R., & Jacobsen, L. (1968). Pygmalion in the classroom. New York: Holt, Rinehart &

Winston.Shellenbarger, S., & Hymowitz, C. (1994, June 13). Over the hill? As population ages, older

workers clash with younger bosses. Wall Street Journal, p. A1.Shore, L. M., Cleveland, J. N., & Goldberg, C. B. (2003). Work attitudes and decisions as a func-

tion of manager and employee age. Journal of Applied Psychology, 99(3), 529–537.Smith, W. J., & Harrington, K. V. (1994). Younger supervisor-older subordinate dyads: A rela-

tionship of cooperation or resistance? Psychological Reports, 74, 803–812.Sopranos, K. (1999, January 24). It’s the age of the younger boss: Sparks may fly when older

workers are asked to report to a younger supervisor. Chicago Tribune, p. 1.Stein, D., Rocco, T. S., & Goldenetz, K. A. (2000). Age and the university workplace: A case study

of remaining, retiring, or returning older workers. Human Resource Development Quarterly,11(1), 61–80.

Tsui, A. S., Egan, T. D., & O’Reilly, C. A. (1992). Being different: Relational demography andorganizational attachment. Administrative Science Quarterly, 37(4), 549–580.

Tsui, A. S., & O’Reilly, C. A. (1989). Beyond simple demographic effects: The importance of relational demography in superior-subordinate dyads. Academy of Management Journal,32(2), 402.

Tsui, A. S., Xin, K. R., & Egan, T. D. (1996). Relational demography: The missing link in verti-cal dyad linkage. In S. L. Jackson & M. N. Ruderman (Eds.), Diversity in work teams: Researchparadigms for a changing workplace (pp. 97–129). Washington, DC: American PsychologicalAssociation.

Venneberg, D. L. (2006). The experience of retirees and their decision to return to the workforce:Implications for organizations. In T. S. Rocco & J.G.L. Thijssen (Eds.), Older workers, new directions:


Employment and development in an ageing labor market (pp. 168–177). Miami: Center for LaborResearch and Studies, Florida International University.

Vroom, V. H. (1964). Work and motivation. New York: Wiley.Waclawski, J. (2002). Large-scale organizational change and performance: An empirical exami-

nation. Human Resource Development Quarterly, 13(3), 289–306.

Mary Hair Collins is professor of business and education at Strayer University OrlandoEast Campus in Orlando, Florida.

Joseph F. Hair, Jr. is professor of marketing at Kennesaw State University, Kennesaw,Georgia.

Tonette S. Rocco is associate professor and program leader of adult education and human resource development in the College of Education at Florida InternationalUniversity, Miami.



The Older-Worker-Younger-Supervisor Dyad: A Test of the Reverse Pygmalion Effect

Documents

Transcript of The Older-Worker-Younger-Supervisor Dyad: A Test of the Reverse Pygmalion Effect