How Salary Communicates: Feedback in
the Context of Major League Baseball
Jon Bruschke, Cal. State Fullerton
John Cunningham, Baylor University
Glenn Abernathy, Jerry Weintraub Productions
This Paper was presented at the 1998 National Communication Association Convention in New York, NY. (Home of the New York Mets).
Motivation is a topic that is of obvious interest to communication scholars and yet is one that has been strangely neglected. As Cusella (1987) notes: "Researchers exploring relationships between incentive/rewards and motivation/performance have not adequately considered the potential effects of the communicative aspects ... In fact, we are aware of no studies that have systematically explored the communication dynamics that surround reward administration" (p. 638). Such a neglect is odd; certainly, whatever else they may do, organizations function to perform certain tasks. Given this orientation motivating employees"to perform well on tasks is quite often the sine qua non of organizational existence and is certainly a primary function of management.
Perhaps the neglect of motivational studies can be traced to a misconception about how communication is defined and what qualifies as a communication study. Cusella's comment demonstrates what is probably the prevailing view about communication; it is something that surrounds and is apart from the reward that is offered. On the other hand, it may be that the reward itself is feedback and serves to convey meaning. Simply put, rewards are not surrounded by communication, rewards are communication. A salary decision is a message. Although communication scholars may just now be arriving at this understanding for years the common language has recognized that "Money talks." The case that rewards are communication will be made in three parts: First, rewards convey meaning, second, they serve as feedback, and third, alternative conceptions of communication are reductionistic and linear in their origins.
First, if communication is defined as the conveyance of meaning, it is arguably the case that the transfer of material resources by an organization is the most important form of communication. Consider a phone caller who contacts a large, bureaucratic health care organization and is put on hold for half an hour during which a recorded voice continually repeats "Your call is important to us ... please wait." While the verbal message overtly states that the call is valued the fact that the organization is unwilling to hire sufficient staff to handle incoming calls communicates in no uncertain terms that organization couldn't care less about the phone call and has all the business that it can handle. A Communication scholar who focused only on the verbal aspect of the communication and ignored the material resource decisions would badly misunderstand the meaning that was being exchanged.
Second, in the particular terms of organizational rewards it is relatively easy to demonstrate that rewards serve as feedback. For the present purposes, salary will be considered as the working example of a reward simply because monetary rewards are the most prevalent way that organizations attempt to acknowledge employee performance. Put in the traditional syllogism, salary is feedback, feedback is communication, and thus salary is communication. Feedback includes the messages conveyed to a receiver about his, her, or its group performance (Cusella, 1987) and salary meets all of the crucial characteristics of feedback. Initially, salary meets all of Cusella's (1980) characteristics of feedback: It can reward (strengthen habits or behavior), motivate (increase the perceived value of successful performance and decrease the perceived value of unsuccessful performance), and regulate (keep goal-directed behaviors on course). In addition, it can provide some information about the correctness, accuracy, or adequacy on an individuals past behavior (Bourne, 1966). Finally, formal organizational rewards have been identified as one of Grellers (1980) six sources of feedback.
Third, Clair and Thompson (1996), addressing pay discrimination in the Journal of Applied Communication Research, have eloquently argued that definitions of communication that exclude pay miss important aspects of the material reward system. Clair and Thompson contend that "pay inequity itself is an articulation of the currently accepted practices in the American socio-economic system ... Pay inequity is an articulation, a discursive practice" (p. 2). In fact, the authors contend that traditional conceptions of pay as something apart from communication are inadequate: "By viewing pay inequity as an outcome these theories tend to conceive of the relationship between communication and pay inequity in a linear fashion and a reductionistic manner" (p. 2). Following Weick's suggestion that simple cause/effect theorizing be avoided, the authors argue against treating salary either as a cause of communication or an outcome of it, but rather seek to reveal how pay and communication "contain each other" (p. 2). The premise set forth in this paper is that salary, like pay equity, should be conceptualized as an articulation or discursive practice.
The present study will extend these conceptions of salary as communicative feedback by studying the linkage between salary and performance. While at present salary is certainly a predominant mode of providing employee feedback little is known about its effectiveness. If we more fully understand the processes that motivate better employee performances it will be possible to devise more effective strategies for enhancing performance. As is so often the case with communication studies, it may very well be that the most obvious approach to altering desired outcomes is not the most effective, efficient, or humane.
Models and Design Rationale
Motivation and Salary
Cusella (1987) has provided a thorough review of current approaches toward motivation. The primary variables of interest are motivation, performance, and ability. Performance is generally thought to be the product of a (perhaps multiplicative) function of motivation and ability. Motivation may be further divided into intrinsic and extrinsic categories. It is plausible to suppose that intrinsic and extrinsic factors lead to motivation and motivation and ability in turn influence performance. Current research supports the thesis that "the generally accepted view is that the impact of feedback on performance is virtually always positive ... [although] the relationship depends on the type of feedback involved" (Cusella, 1987, p. 638). In other words, feedback does relate to performance, and the type of feedback matters.
Salary may be integrated into the current understandings of motivation theory in two ways. First, salary is an extrinsic source of motivation. Thus, if performance is indeed a function of ability, intrinsic, and extrinsic motivation, salary might only be expected to have modest effects on performance since it is only one of three theoretically relevant variables. This framework suggests the need for comparative studies of different types of motivation. Thus the central research question addressed here is: Are salary levels related to performance, and if they are, how strong is the relationship?
Second, because the type of feedback seems relevant to the relationship between performance and motivation exploring the different ways that salary might function as feedback are important. In the context of salary, one possible distinction is between that of rewarding past behavior, in which case the feedback is retrospective, and offering incentives for future behavior, in which case the feedback is prospective. In the case of retrospective feedback, the message essentially is that "You have done well in your past behavior." A prospective approach sends the message, "If you do well you will be rewarded."
The prospective approach will be labeled the Inducement model, and the retrospective approach will be labeled the Reward model. Both models essentially adhere to the principle that salary motivates performance; however, they differ in the ways that salary and performance are linked. The Inducement model assumes that salary motivates performance in a prospective way so that the possibility of future salary drives future performance. An inducement approach seeks to motivate by offering relatively immediate rewards for good performance; an incentive-based contract or a short-term salary jump would be offered in exchange for an enhanced performance. Salary and performance are thought to be linked to each other in a relatively short period of time. The Reward model, on the other hand, assumes that salary-givers use past performance to fix the level of present salary. The time distance between the performance and the reward is relatively large. An inducement approach seeks to motivate by linking specific performances to specific increases, whereas a reward approach seeks to motivate with the certainty of large but relatively less specific rewards for consistent but generally unparticularized performance.
These models are largely heuristic and are not necessarily mutually exclusive. They are intended to summarize a package of salary-offering strategies. It is not yet necessary to predict which approach will better stimulate improved performance, but a second research question is: Do Reward as compared to Inducement approaches relate to performance differently?
Considerations of Context
In addition to the absence of communication research on motivation in general, Cusella (1987) lamented that: "From a communication theory perspective, one gap in the nonmediated feedback research, in general, is the effect that context has on the feedback performance/motivation relationship" (p. 641). There are a number of different ways to incorporate notions of context into communication studies. One approach is to manipulate the variables experimentally in a laboratory setting. These attempts are limited, however, by the very nature of context. Because context includes all the various factors surrounding and encompassing communication any attempt to isolate individual factors, while well-intentioned, is arduous to achieve and fraught with difficulty at best. Context by its very nature is a multiplicity of interacting factors, and attempts to control and isolate context may prove no more fruitful than efforts to contain and isolate the various parts of a hurricane.
An alternative is to study the variables of interest in naturally occurring contexts. This approach has the additional advantage of enhancing external validity (Anderson, 1987, pp. 120-121). Knowledge about context can accumulate if researchers, when conducting in-context studies, define the parameters of those contexts so that studies of similar variables in different contexts may be compared. If, for example, studies of salary and motivation produce different results in manufacturing as opposed to entertainment industries, scholars may compare the two different industries to develop theories about what differences in context account for the differences in results.
The context selected for the current study is Major League Baseball. Baseball is a useful context for the present study for several reasons. First, there are well known and easy to obtain measures of performance (Kahn, 1993). A batter can be judged by batting average, home runs or runs batted in, and a pitcher by earned run average, wins, or strike outs. Secondly, salary is certainly an issue to most players as witnessed by the numerous player holdouts and the recent strike. In addition, salary figures are relatively easy to obtain and, since they are usually a matter of the public record, do not involve compromising privacy. Finally, baseball is a well-studied academic area. Scholars from various fields have examined the relationship between age and performance (Schulz et al., 1994), performance over time (Hoffman, Jacobs & Gerras, 1992) and in pressure situations (Davis & Harvey, 1992), how eye color and race relate to performance (Beer & Beer, 1989), how playing surfaces relate to performance (Goodman & McAndrew, 1993) and even how later life tragedy is linked to performance (Lester & Topp, 1989).
Of particular interest are those studies that have linked broader theories of motivation or management to player performance. Kahn (1993) investigated the impact of quality of management on individual performance and found that players tend to play better as the quality of management increases. Psychological studies have utilized Equity theory (see, e.g., Adams, 1965; Walster, Walster, & Berscheid, 1978) and Expectancy theory (see, e.g., Vroom, 1964; Porter & Lawler, 1968; Mitchell & Biglan, 1971; Nadler & Lawler, 1977) to explain variance in baseball players' performance. Harder (1991) applied the Equity and Expectancy theories to free agents with the assumption that free agents tend to feel under rewarded. He found that for free agents, as compared to players with signed contracts, home run ratios (which are considered to be more important when deciding salary) increased. Harder (1992) later suggested that free agents, who realize their performance will directly relate to salary for the next season, have less of a performance drop-off. These studies suggest that there may be a theoretically explainable relationship between salary, motivation, and performance in the area of Major League Baseball although to date no such comprehensive study has been conducted.
There are some features of the baseball context that should be noted from the outset. Performance in baseball is obviously related to ability; in fact, all the players at the major league level have been extensively screened and coached. Second, intrinsic rewards are highly relevant to the players. Players are increasingly expressing a willingness to play for a team that will be a winner regardless of salary. Third, careers tend to be short but salary rewards are astoundingly high and thus there is an emphasis for players to earn as much as they can as quickly as they can. Finally, baseball is a very public context in which player performance is scrutinized by management, other players, scouts, opposition, media and fans. With these features of the context in mind, baseball will serve as the site for this study of salary and motivation.
Variables and Operations
Overview of Analysis Strategy
Baseball is a useful arena for study because each reward model seems to be implicit in the salary setting procedures. The use of an incentive clause in a contract is the clearest example of the inducement model. If a player has an incentive clause in his contract some performance standard is set, such as a minimum number of home runs or pitching a certain number of innings, and if the player attains the performance standard there is a set monetary reward. Similarly, a large salary or a jump in salary might induce better player performance. Alternatively, salary may be used as a reward for past performance and thus is indicative of the reward model. If the reward model more accurately describes the way that salary is offered by management, salary is awarded largely on the basis of past performance. For example, if salary was set based on a reward system a player that led the league in home runs in 1991 could expect a large salary in 1992 regardless of 1992 performance.
To test the differing models the same data set was analyzed twice. To test the Inducement model, salary variables were treated as independent variables and performance variables were treated as dependent variables. To test the Reward model the assumptions were reversed, and performance variables were treated as independent variables whereas salary variables were treated as dependent variables. If the Inducement model better explains the data then salary will drive performance and not the opposite. If the Reward model better fits the data then performance will drive salary and not the reverse.
Data were collected for three years covering the seasons 1991-1993. The time-ordered nature of the data raises two issues. First, because several of the variables described below were time lagged (for example, a jump in salary must be measured by subtracting salary in one year from the salary in the preceding year) statistical tests could only be conducted on the 1992 and 1993 data. A second issue, described by Sayrs (1989), is that researchers dealing with time-series data must decide either to pool the data under the assumption of equivalence of effects across time or, instead, treat each time panel separately. In the present case, pooling would mean collapsing data for the 1992 and 1993 seasons. Because there was more than one performance variable for batters and three performance variables overall, and because the independent variables in one model served as dependent variables for the other model, for the sake of simplicity of analysis the data were not pooled. As a result, separate tests were conducted for batters and pitchers and for different years. The analysis of the Inducement model, for example, involved separate tests for pitchers and batters, and for both pitchers and batters different tests were conducted for the 1992 and 1993 seasons. While this approach sacrifices a single test for significance, the research decision simplified the analysis. Furthermore, the overall pattern of results should provide a meaningful interpretation of the data.
The Inducement Model
Five different salary variables operationalized the independent variables for the Inducement model; the performance variables described in the following section represented the dependent variables. The variable "Salary" represents a player's salary for a given year. "Salary-1" represents the player's salary in the preceding year. "Salary Jump" represents the increase in salary between a present year and the immediately preceding year. Finally, the salary structure of baseball recognizes seniority and there is a strong relationship between a player's tenure in Major League Baseball and his pay, both as a recognition of service and because only fairly good players can have long careers. A player may not feel rewarded by absolute salary numbers but rather by his pay in relation to his length of time in he league. Thus, "Relative Salary" was computed by subtracting predicted scores based on a player's tenure in the league from the player's actual salary (because of the slightly curvilinear nature of the relationship quadratic correlations were computed for all players for all years. Significant results were obtained for every year examined; in all cases quadratic functions explained more variance than simple linear functions). Positive numbers represent players who are relatively well paid in relation to equally tenured players and negative numbers represent players who are paid relatively poorly. The variable "Incentive" represents a dollar figure that was part of the annual salary that was linked to specific performances.
The Reward Model
Performance variables represented the independent variables for the Reward model and the preceding salary variables represented the dependent variables. Because pitchers and hitters do essentially different things there are different measures of performance for each position. For pitchers, an Earned Run Average or "ERA" represents the average number of earned runs the pitcher allowed over a nine-inning span. This well-accepted measure of a pitcher's quality was used to measure pitcher performance. For batters, batting average (hits divided by at-bats) was used as one performance measure (designated "BAvg"). Batting average is among the most popular and yet most criticized measures of hitting performance, however, and thus the number of runs-batted-in per at-bat (designated "RBI/AB") was used as a second measure of batting performance. Because performance may level off the relative increase or decline in a player's performance may be of special interest, and thus separate measures of player improvement were constructed for both RBI/AB and BAvg. "RBI/AB Jump" represents a players' RBI/AB in a current year subtracted from the previous year, and "BAvg Jump" is the BAvg for a current year subtracted from the same variable in the previous year. Because tenure in the league can represent, at least globally, the sum total of a player's past performance quality, "Years" was retained as a Reward variable and represents the number of years a player has been in the Major League Baseball.
Finally, one variable was included that overlapped the two models. Under the terms of the collective bargaining agreement in place for the years from which the data were gathered, a player could not become a free agent until the completion of six years of major league service. Because of the lack of leverage in bargaining negotiations, players who had not yet attained free agent status might have, on average, lower salaries. Players in their free agent year (their sixth in the league) might tend to be especially motivated under Reward conditions because good performance in the free agent year would likely result in higher bidding for the player as a free agent. On the other hand, the motivation does resemble an Inducement motivation to a large degree; free agent status might be thought of as a large incentive clause with an unspecified but certainly large payoff. The variable "Free Agent" was dummy coded as "1" for players in their free agent year and "0" for all other players and represents a cross-over between the two models.
Analyses and Results
Tests of the Inducement Model
Batters. A multiple regression analysis was conducted for both the 1992 (n=202) and 1993 (n=189) seasons. A mixed stepwise procedure was used with a probability to enter and exit set at .25. Separate analyses were conducted for each of the four performance variables. Because relatively few players had incentive clauses in their contracts, including the Incentive variable in the overall analysis resulted in excluding over three-fourths of the data, and thus it was tested separately. For all other analyses, all four Inducement variables (Salary, Salary-1, Salary Jump, and Relative Salary) were included along with the mixed variable (Free Agent).
Insert Table 1 about here
The Incentive clause variable did not significantly correlate with any of the four performance variables for either 1992 or 1993 data. The results of the eight regression equations appear in Table 1. As the Table reveals, the only result that is anything but trivial is the relation between Relative Salary and RBI/AB. In general, a higher relative salary is related to a higher RBI/AB percentage; in 1992 14% of the variance accounted for; the figure was a more modest 6% for 1993.
Pitchers. Similar regression procedures were used for pitchers for both 1992 (n=133) and 1993 (n=136). All five independent variables were again included, and two dependent variables (ERA and ERA Jump) were used. As with batters, the Incentive variable was tested separately and failed to correlate significantly with either dependent variable for either year. No variable was entered into either the ERA or the ERA Jump equation for 1992 or into the ERA Jump equation for 1993. For the 1993 ERA equation, Relative Salary was entered at the .0584 significance level and accounted for an adjusted R2 of .03. These results can only be interpreted to mean that for the pitchers, the Inducement model was of almost no use in predicting ERA performance.
The Reward Model
Batters. Three Reward variables (Years, RBI/AB for the previous year, and BAvg for the previous year) were entered into a regression equation with Salary as the dependent variable for both 1992 and 1993. Regression procedures were identical to those for the Inducement tests. Table 2 summarizes the results. As the Table reveals, the Reward variables were all significantly related to Salary, although the combined effect was much stronger for 1992 than for 1993.
Insert Table 2 about here
Pitchers. Three Reward variables (Years, ERA-1, and ERA Jump) were entered into a multiple regression equation for 1993; criteria for entry and exit in the stepwise procedure were identical to the other analyses. Salary measures were treated as the dependent variable. Because no ERA Jumps were possible for 1992 (the 1990 to 1991 calculation was not possible because no 1990 data were collected) the ERA Jump variable was excluded for 1992 and the procedures for analysis were otherwise identical to the regression analysis for batters. Table 3 summarizes the results. The results tend to mirror those for the batters, so that the combined Reward variables are strongly related to salary and the effect is larger for 1992 as compared to 1993. The Years variable continues to dominate the variance accounted for, and the remaining variables account for much less variance for pitchers as compared to batters. Overall, the effect size is somewhat smaller for pitcher than for batters.
Insert Table 3 about here
Summary of Results
Overall, the Reward model fared much better than the Inducement model. When salary is treated as an independent variable there are weak relationships with performance. However, when the equations are reversed and salary is treated as the dependent variable and performance is the independent variable, very creditable statistical relations emerge. Overall, performance accounts for somewhere between 16 and 46 percent of the variance in salary, although the relationship seems fairly unstable across time and is somewhat stronger on average for batters than for pitchers. Within the performance variables, the length of tenure in the league is most strongly associated with salary so that Years alone accounts for as much variance in salary as the other performance variables combined. In sum, the current salary system in Major League Baseball is based on a Reward system, so that length of service in the league is the primary determinant of salary while other performance variables account for the remaining variance.
There were some other intriguing elements of the findings. The Inducement model did produce a significant relationship between Relative Salary and RBI/AB for both years with a magnitude of about 13 to 14%. Furthermore, the Inducement model fared better with batters than with pitchers. Although it fared well, the Reward model at its best accounted for less than half the variance in salary. The Reward model fared starkly better for pitchers in 1992 than 1993. Finally, if the effect for years is removed from the Reward model, it fares little better than the Inducement model with RBI/AB treated as the dependent variable.
Discussion and Implications
The most significant finding was that the Reward model seems to fit the data much better than the Inducement model, a conclusion that answers the second research question. To the extent that the data are accurately measuring the concepts under consideration, this means that the current system of reward in major league baseball is one that promises large future rewards to players who can obtain a high performance level or survive in the league for several years. Perhaps as importantly, specific inducements do not appear to correlate with specific performance. These findings contrast those of Harder (1992).
In relation to the first research question, this means that extrinsic motivations are relatively unimportant to performance in major league baseball. By default, this suggests that intrinsic motivation and ability are more crucial to performance. Theoretically, at least in contexts in which ability is carefully screened and intrinsic motivation may be high, it appears that extrinsic motivations are relatively unimportant. This finding is important because it suggests that, in contexts where intrinsic motivation and ability are particularly salient, extrinsic motivation is not merely of reduced importance but almost vanishes as a contributing factor entirely.
The practical implications of these conclusions are possibly more significant than the theoretical insights. While the feedback salary provides seems to be relatively unimportant to performance, Kahn (1993) discovered that management quality was significantly related to performance. This suggests that baseball teams seeking to enhance performance should not focus solely on salary incentives to remain competitive. Baseball's recent strike focused almost exclusively on salary issues, and as a related topic, the ability of small-market clubs to compete against richer counterparts (Chass, 1994; "What the Fuss is About," 1994). At least in relation to the small-market issue, it seems that small market teams could enhance their competitiveness by compensating for a lack of salary incentives by hiring a better manager. Indeed, there is no significant correlation between market size and the summed wins for major league teams over the past 15 years (df=22, p = .17; expansion teams, Canadian teams and 1981 excluded). In an era where the perceived need for a salary cap caused a strike that threatened the existence of the game, alternative communication practices in the form of better management may have had the ability to alleviate the concerns of small market clubs in a way that could have avoided the economic confrontation altogether.
A second practical consideration is that short-term inducements seem to have little to do with specific, short-term performances. This finding suggests that the current and very expensive practice of signing big-name and big-money free agents for a single year in the hopes of winning a pennant may be largely misplaced. Rather than focus on the short-term benefits of signing a single player in a single year, teams would do well to exert their efforts toward the longer-term end of building a strong management system.
As with any individual study the conclusions here are subject to a number of restrictions and are offered tentatively. First, the performance variables may have been subject to a good deal of error. ERA is widely accepted as the clearest measure of a pitcher's performance but is notoriously unstable. Batting average, while it is easily interpretable and accounts well for differences in at-bats, has been widely criticized (Rapaport, 1994). There were no measures of defense or a total number of runs scored. As a result, perennial all-star defensive players such as Ozzie Smith may have been ranked as unproductive with the current set of measurements. It is interesting to note that RBIs per at-bat, arguably the best performance variable utilized, do seem to fit the Inducement model almost as well as the performance variables fit Reward model, with each accounting for roughly 15% of the variance when the Year variable is excluded from Reward calculations. This suggests that tighter measurement of performance might alter any conclusion about the predominance of the Reward model.
Second, the free agent measure may have misidentified a number of players. It was assumed that all players in their sixth year were free agents but at least some players were signed to longer-term deals and thus were not eligible for free agency even though it was their sixth year. Data that identified exactly a player's free agent year would better capture the concept.
Third, the best predictor overall was Years, but since the collective bargaining agreement that binds player salaries includes a sliding scale for the minimum player salary based on years of service the finding is entirely unsurprising. It isn't a meaningless finding because few players survive many years if they aren't very good, but the research could be better if the minimum salary for each year in the league were entered instead of estimating the figure based on a correlation. In addition, since the Years variable was clearly the most substantial variable and essentially operated as a simple control, caution must be exercised when interpreting the meaning of the overall variance accounted for. If Years is not thought to be theoretically relevant the magnitude of the effects may be overstated.
Finally, the results varied greatly across time panels and for pitchers and batters. The overall performance of the Reward model variables dropped by roughly 20 points for both pitchers and batters between 1992 and 1993, and they varied greatly between batters and pitchers for both years. Before reaching any firm conclusions more stable results must be obtained.
The first and most obvious direction for future research would be a replication of this study that remedies some of the limitations of the data set identified above. Different performance measures, more accurate free agent data, and minimum salary levels could all be incorporated into a new analysis. In addition, motivation research should expand to other naturalistic settings to allow for cross-context comparisons that would further aid theory development. The findings here are certainly interesting but baseball is in many ways a unique context. Research that can identify how far the results here extend to other situations would certainly be worthwhile.
Finally, while salary figures are relatively easy to obtain for baseball as opposed to those for other individuals in other professions, it is difficult to tell whether the players are motivated by the public celebrity status, the certainty of salary rewards for career-length performances, or by an intrinsic desire to do well. While the present study did have relatively precise measurements of performance and extrinsic motivation as operationalized by salary, intrinsic motivation and ability were not measured at all. Research that measures all variables under consideration will facilitate more precise theories that can explain the exact relationships between the variables.
In sum, the present study has argued for an expansion of current conceptions of motivation with the argument that salary communicates and should receive attention from communication scholars. Empirical data were gathered from the baseball context and the results revealed that rewards tend to be given in a retrospective as opposed to prospective fashion. There was almost no relation between year-to-year variations in salary and year-to-year variations in performance. This study has offered data relevant to theoretical understandings of motivation and can claim practical significance. It is, however, only a first step toward a deeper understanding of how salary communicates and how it interacts with motivation. The findings are promising enough, however, to suggest that great progress in these areas is possible.
Inducement Model Multiple Regression Results for Batting Data
Year Variable Variable Probability R2
93 BAvg Relative Salary .0052 .04
Salary .0798 .02
Overall R2 adj: .05
93 RBI/AB Relative Salary .0007 .06
Salary Jump .0076 .04
Free Agent .0016 .05
Overall R2 adj: .13
93 RBI/AB Jump Free Agent .1599 .01
Salary Jump .2293 .007
Overall R2 adj: .008
93 BAvg Jump Salary .0457 .02
Overall R2 adj: .02
92 BAvg Relative Salary .0044 .04
Free Agent .1040 .01
Salary Jump .1816 .008
Overall R2 adj: .05
92 RBI/AB Relative Salary .0001 .14
Salary .1194 .01
Overall R2 adj: .14
92 RBI/AB Jump Salary Jump .0025 .04
Overall R2 adj: .04
92 BAvg Jump Salary Jump .0017 .05
Free Agent .1420 .01
Salary .2426 .006
Overall R2 adj: .05
Reward Model Multiple Regression Results for Batting Data
Year Variable Variable Probability R2
92 Salary Years .0001 .25
RBI/AB-1 .0001 .16
BAvg-1 .0001 .05
Overall R2 adj: .46
93 Salary Years .0001 .13
Bavg-1 .0001 .11
RBI/AB-1 .0004 .05
Overall R2 adj: .27
Reward Model Multiple Regression Results for Pitching Data
Year Variable Variable Probability R2
92 Salary Years .0001 .32
ERA-1 .0151 .03
Overall R2 adj: .34
93 Salary Years .0001 .11
ERA-1 .0019 .06
ERA Jump .1258 .01
Overall R2 adj: .16
Anderson, J. A. (1987). Communication research: Issues and Methods. New York, NY: McGraw-Hill Book Company.
Adams, J. S. (1965). Inequity in social exchange. In L. Berkowitz (Ed.), Advances in Experimental Social Psychology (267-299). San Diego, CA: Academic Press.
Beer, J., & Beer, J. (1989). Relationship of eye color to professional baseball players batting statistics given on bubblegum cards. Perceptual and Motor Skills, 69, 631-634.
Bourne, L. E. Jr. (1966). Comments on Prof. I.M. Bilodeaus paper. In E. A. Bilodeau (Ed.), Acquisition of skill (pp. 297-313). New York: Academic Press.
Chass, M. (1994, June 7). Owners pull so hard on cap, They can't see strike in their path. New York Times, p. B11.
Clair, R. & Thompson, K. (1996). Pay discrimination as a discursive and material practice: A case concerning extended housework. Journal of Applied Communication Research, 24, 1-20.
Cusella, L. P. (1980). The effects of feedback on intrinsic motivation. In D. Nimmo (Ed.), Communication Yearbook 4, (pp. 367-387). New Brunswick, NJ: Transaction.
Cusella, L. P. (1987). Feedback, motivation and performance. In F. M. Jablin, L. L. Putnam, K. H. Roberts, & L. W. Porter (eds.), Handbook of Organizational Communication: An Interdisciplinary Perspective (pp. 624-678). Newbury Park, CA: Sage.
Davis, M. H., & Harvey, J. C. (1992). Declines in major league batting performance as a function of game pressure: A drive theory analysis. Journal of Applied Social Psychology, 22(9), 714-735.
Goodman, G. H., & McAndrew, F. T. (1993). Domes and astroturf: A note on the relationship between the physical environment and the performance of major league baseball players. Environment and Behavior, 25(1), 121-125.
Greller, M. M., (1980). Evaluation of feedback sources as a function of role and organizational level. Journal of Applied Psychology, 65, 24-27.
Harder, J. W. (1991). Equity theory versus expectancy theory: The case of major league baseball free agents. Journal of Applied Psychology, 76(3), 458-464.
Harder, J. W. (1992). Play for pay: Effects of inequity in a pay for performance context. Administrative Science Quarterly, 37, 321-335.
Hoffman, D. A., Jacobs, R., & Gerras, S. J. (1992). Mapping individual performance over time. Journal of Applied Psychology, 77(2), 185-195.
Kahn, L. M. (1993). Managerial quality, team success, and individual player performance in major league baseball. Industrial and Labor Relations Review, 46(3), 531-547.
Lester, D., & Topp, R. (1989). Major league baseball performances of players who were later suicides or homicide victims. Perceptual and Motor Skills, 69(1), 272.
Mitchell, T. R., & Biglan, A. (1971). Instrumental theories: Current uses in psychology. Psychology Bulletin, 76, 432-454.
Nadler, D. A., & Lawler, E. E. (1977). Motivation: A diagnostic approach. In J. R. Hackman, E. E. Lawler, & L. W. Porter (Eds.), Perspectives on Behavior in Organizations (pp. 26-34). New York, NY: McGraw-Hill.
Porter, L. W., & Lawler, E. E. (1968). Managerial Attitudes and Performance. Homewood, IL: Irwin.
Rapaport, R. (1994, April 11). Stat's all, folks. Forbes, pp.S62-S71.
Sayrs, L. W. (1989). Pooled time series analysis (Series: Quantitative applications in the social sciences). Newbury Park, CA: Sage.
Schulz, R., Musa, D., Staszewski, J., & Siegler, R. S. (1994). The relationship between age and major league baseball performance: Implications for development. Psychology and Aging, 9(2), 274-286.
Vroom, V. H. (1964). Work and Motivation. New York: Wiley.
Walster, E., Walster, G. W., & Berscheid, E. (1978). Equity: Theory and Research. Boston, MA: Allyn & Bacon.
What the fuss is all about. (1994, July 29). New York Times, p. B9.