Give Them a Fishing Rod, if It Is Not Urgent: The Impact of Help Type on Support for Helpers’ Leadership PDF

Journal of Applied Psychology © 2023 American Psychological Association 2024, Vol. 1...

Journal of Applied Psychology © 2023 American Psychological Association 2024, Vol. 109, No. 4, 551–572 ISSN: 0021-9010 https://doi.org/10.1037/apl0001155 Give Them a Fishing Rod, if It Is Not Urgent: The Impact of Help Type on Support for Helpers’ Leadership Lily Chernyak-Hai1, Daniel Heller2, Ilanit SimanTov-Nachlieli2, and Merav Weiss-Sidi2 1 School of Business Administration, Peres Academic Center 2 Coller School of Management, Tel Aviv University Taking a follower’s perspective on leadership and contributing to the new research stream on behaviors conducive to its emergence, we examined how distinct types of instrumental (task focused) helping— This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. autonomy- versus dependency-helping—affected recipients’ support for their helpers’ leadership. Based on This document is copyrighted by the American Psychological Association or one of its allied publishers. the literature on employees’ needs for autonomy and mastery, combined with the empowering nature of autonomy-helping, we reasoned that autonomy- (vs. dependency-) helping typically signals greater benevolence toward recipients, enhancing their support for their helpers’ leadership. Our findings were generalized across various samples (of U.S. and Israeli employees), manipulations, and research settings: simulations (Studies 1 and 2b), workplace role-play scenario (Study 2a), and recollections of helping events in the workplace (Study 3). We found that autonomy- (vs. dependency-) helping increased recipients’ support for their helpers’ leadership by heightening perceptions of helpers’ benevolence-based (rather than ability-based) trustworthiness (Studies 1 and 3). We also showed time pressure to be a boundary condition under which the advantage of autonomy-helping disappeared (Studies 2a and 2b)—with dependency- helping then inducing comparable levels of perceived benevolence and thus similar support for the helper’s potential leadership. Overall, we shed light on the development of informal leadership by uncovering how recipients interpret and respond to the two help types. Practically, this analysis opens the door to new ways for aspiring managers to enhance support for their leadership from potential followers, available even to those unlikely to be appointed to formal leadership positions. Keywords: leadership support, perceived benevolence, trust, autonomy- and dependency-oriented helping, informal leadership Supplemental materials: https://doi.org/10.1037/apl0001155.supp Consider an employee asking a colleague to help them with a work- which help type is more beneficial for gaining recipients’ support for related problem. In one scenario, the colleague provides the employee their helpers’ leadership, why, and when. with the full solution to the problem at hand, leaving the employee In modern-day organizations, which are increasingly switching dependent on the colleague’s assistance the next time a similar to flat structures in which leadership shifts dynamically, leaders problem arises. In the second scenario, the colleague provides tools increasingly emerge informally (Bedeian & Hunt, 2006; Howell & and explanations for solving the problem autonomously. These two Shamir, 2005). This trend renders follower’s leadership support— different instrumental approaches to helping are termed dependency- defined as the degree to which followers are willing to grant and autonomy-oriented helping, respectively (Nadler, 1998, 2002). leadership (Platow & van Knippenberg, 2001; Ullrich et al., 2009)— Here we examine whether help recipients differentiate between such cardinal in the development of informal leadership (DeRue & distinct help types, what causal inferences they make about the Ashford, 2010). That is, it is a precursor, and often a prerequisite, to helper’s behavior (Kelley, 1973), and how they evaluate their helper leadership emergence (Epitropaki et al., 2013; Howell & Shamir, in each case. Addressing the broader critical and increasingly relevant 2005; Kalish & Luria, 2016). question of how informal leadership develops (e.g., Marinova et al., In this article, we unpack the largely neglected psychological 2013; McClean et al., 2018; Van Kleef et al., 2021), we aim to reveal process whereby followers grant informal leadership to others This article was published Online First November 16, 2023. is alphabetical. Ilanit SimanTov-Nachlieli https://orcid.org/0000-0002-2713-7686 Lily Chernyak-Hai played an equal role in writing–original draft and This research was supported by the Israel Science Foundation under Grant writing–review and editing. Daniel Heller played an equal role in writing– 2697/20 awarded to Daniel Heller and Ilanit SimanTov-Nachlieli and by The original draft and writing–review and editing. Ilanit SimanTov-Nachlieli Henry Crown Institute of Business Research in Israel awarded to Ilanit played an equal role in writing–original draft and writing–review and editing. SimanTov-Nachlieli. Merav Weiss-Sidi played a supporting role in writing–original draft and Merav Weiss-Sidi is the postdoctoral fellow under the supervision of writing–review and editing. Daniel Heller and Ilanit SimanTov-Nachlieli. Correspondence concerning this article should be addressed to Ilanit Lily Chernyak-Hai, Daniel Heller, and Ilanit SimanTov-Nachlieli have SimanTov-Nachlieli, Coller School of Management, Tel Aviv University, contributed equally to the production of this article. Order of authors’ names P.O. Box 39040, Tel Aviv 6997801, Israel. Email: [email protected] 551  552 CHERNYAK-HAI ET AL. (Badura et al., 2022; Gan et al., 2018; Hanna et al., 2021; cf. Van and innovation; Podsakoff et al., 2009), helping behaviors also benefit Kleef et al., 2012, 2021). We focus on follower-granting leadership helpers by potentially facilitating their ascension to leadership. For based on receiving transient instrumental helping behavior (e.g., example, correlational research has shown that employees who are Flynn, 2003; Marinova et al., 2013), defined as task-focused helpful to more of their team members, but who seek help from fewer, behavior aimed at assisting to complete or resolve a work task or are conferred higher status (Flynn, 2003; Flynn et al., 2006). More problem (Geller & Bamberger, 2009; Settoon & Mossholder, 2002). directly, helping behavior and leadership emergence are positively Yet, as evident from the opening example, there are two correlated (Marinova et al., 2013). However, the effect of help type on fundamentally distinct types of instrumental helping. Going beyond recipients’ support of a potential leadership role for the helper has the role of helping in general and based on the pivotal role of trust in not previously been examined. Filling this gap to enhance our leadership emergence (Marinova et al., 2013), we posit that the help understanding of how informal leadership develops, we distinguish type plays an important role in this process by driving different here between two basic types of instrumental helping that we expect to perceptions of the benevolence of helpers. Drawing from theorizing invoke disparate levels of leadership support. on employees’ intrinsic need for autonomy (Deci et al., 2017; Instrumental helping of a person in need facing a current, This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Hackman & Oldham, 1975; Herzberg, 1966), in conjunction with potentially recurring work task (e.g., completing a periodic report) This document is copyrighted by the American Psychological Association or one of its allied publishers. the long-term empowering nature of autonomy-oriented helping can vary considerably in content, temporal relevance, and needs (Nadler, 2020), we predict that such helping (vs. dependency- addressed. Dependency-helping—providing a final solution to the oriented helping; hereafter referred to as autonomy- and dependency- problem at hand—offers immediate, short-term relief but also helping) typically signals greater benevolence toward recipients, thus reinforces recipients’ dependency on the helper or others for future increasing their support for their helpers’ leadership. Finally, we assistance. In contrast, autonomy-helping provides the recipients offer a situational contingency that this advantage of autonomy- with tools and explanations necessary to resolve the current and helping is eliminated in time-pressured tasks. similar future problems autonomously; however, it does not offer This research offers several theoretical contributions to the immediate relief, and it requires additional effort by the recipients. leadership and helping literatures. First, given the increasing This temporal trade-off in benefits is also evident in the different centrality of informal leadership in organizations (Bedeian & Hunt, needs satisfied by the two types: Whereas dependency-helping 2006; Howell & Shamir, 2005), it is vital to understand its proximal fulfills a concrete performance need, autonomy-helping satisfies behavioral antecedents (Lanaj & Hollenbeck, 2015; McClean et al., recipients’ long-term needs for autonomy and mastery (Nadler, 2018). Further, given the inherently social nature of informal 1998, 2002, 2020). leadership, as rooted in followers’ volition (Badura et al., 2022; To date, the study of these two pervasive types of helping in DeRue & Ashford, 2010; Hanna et al., 2021), we believe that organizations has been surprisingly limited. Notable exceptions understanding the mediating role of follower perceptions in this include research on the antecedents and outcomes of help-seeking process is critical (Badura et al., 2022; Hanna et al., 2021). Going (Bamberger, 2009; Geller & Bamberger, 2012; Liu et al., 2022). beyond the focus of prior research on perceived competence in More directly relevant is recent research on help giving and leadership emergence (Anderson & Kilduff, 2009; DeRue et al., specifically on antecedents of helpers’ choice to provide one or the 2015; Judge et al., 2004), here we illuminate the role of perceived other type of help. For example, helpers were more likely to opt for benevolence in leadership emergence. dependency-helping when asked to pick one of the two help types Second, we go beyond previous research on help type in for their envied peers (Montal-Rosenberg & Moran, 2022), as were organizations, which has mostly focused on the helper’s perspective helpers in the competitive context of equity (vs. equality) pay (e.g., Bamberger & Levi, 2009; Montal-Rosenberg & Moran, 2022) structures (Bamberger & Levi, 2009). Similarly, theorizing in the or on the role of helping in general (e.g., Flynn, 2003; Marinova et intergroup domain points to dominant groups’ preference to give al., 2013) by taking the perspective of recipients (followers). That is, dependency-helping to low-status groups to maintain their dominant we advance the novel proposition that when evaluating helpers, position (Nadler, 2002). Although those are clearly not the only recipients who are motivated to satisfy intrinsic psychological needs motives underlying helpers’ choice to offer dependency- (vs. for autonomy and mastery, beyond extrinsic needs or immediate autonomy-) helping, these studies are suggestive of potential less performance goals (Deci et al., 2017; Hackman & Oldham, 1975; benevolent (or more selfish) motivations driving that choice. Herzberg, 1966), interpret and respond differently to distinct types However, how the two help types are viewed or interpreted by of instrumental helping. We thereby challenge the implicit assumption recipients, who are typically not privy to their helpers’ underlying that instrumental helping, regardless of type, yields uniform outcomes, motivations, needed to be examined. arguing that the two help types have distinct and far-reaching workplace implications. Practically, our research suggests new ways Help Type and Benevolence-Based Trustworthiness for aspiring managers to enhance support for their informal leadership, available even to those who are unlikely to be appointed to formal “Give a man a fish, and he’ll eat for a day. Teach a man to fish, and leadership positions. he’ll eat for a lifetime.” As noted in this ancient proverb, given its long-term benefits, autonomy-helping is typically theorized to be more beneficial to recipients than dependency-helping (Maimonides, Autonomy- and Dependency-Helping 1998; Nadler, 2020). However, somewhat surprisingly, it is currently Helping in the workplace is an important prosocial behavior unknown how recipients themselves interpret and respond to each (e.g., Bolino & Grant, 2016; Hofmann et al., 2009) and a key element of help type, and how they evaluate those providing them. Thus, we organizational citizenship behavior (OCB; Organ, 1988, 1997). Besides explore the implications of the help types for recipients’ perceptions yielding positive outcomes for help recipients (e.g., productivity of helpers’ benevolence-based trustworthiness, defined as the extent  AUTONOMY-HELPING, DEPENDENCY-HELPING AND LEADERSHIP SUPPORT 553 to which an actor is believed to want to do good for the trusting party, importance of followers’ perceptions of benevolent intentions aside from an egocentric profit motive (Mayer et al., 1995). Given toward them in granting leadership in the first place. how challenging and biased social perceptions of actors’ motivations According to Kramer (2011), followers’ trust in prospective can be (Epley, 2008) and that attribution of benevolent motives for leaders is a function of two primary perceptions: (a) perceived helping can influence judgments of the helper’s character, as well as competence—confidence in their competency to perform work tasks reactions toward them (Berman & Silver, 2022; Bowler et al., 2019; at a high level and (b) perceived benevolence—perception of their Inesi et al., 2021), this question needs close, empirical scrutiny. intention to care for them. Though related, these two perceptions are Given the inherently instrumental nature of dependency-helping— distinct because the trustor’s confidence in the trustee’s abilities and providing recipients with the full solution (and thus an immediate in their intentions toward the trustor (benevolent or not) are separate. relief in regard) to a problem or task at hand—we posit that those Further, whereas the first perception is character-based or cognitive, providing such help (compared to no help) are perceived favorably by the latter is affective and specific to the particular follower– recipients, meaning that dependency-helping signals greater benevo- prospective leader relationship. lence than no help. Indeed, OCB and helping behaviors are typically Though the role of perceived competence in leadership emergence This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. viewed as positive, signaling concern for the recipients’ welfare has been addressed (e.g., Anderson & Kilduff, 2009; DeRue et al., This document is copyrighted by the American Psychological Association or one of its allied publishers. (McAllister, 1995; Moorman et al., 2018). 2015; Judge et al., 2004), the influence of perceived benevolence is a Nonetheless, we argue that autonomy-helping typically signals more novel notion. This has been a conspicuous oversight because even greater benevolence. In developing this argument, we draw on this is likely a critical antecedent of discretionary leadership support canonical research on intrinsic motivation showing that employees that is important in mitigating followers’ risk perceptions. That is, are inherently motivated to fulfill basic, intrinsic needs for autonomy perceived benevolence enables the recipient’s belief that such and mastery, beyond extrinsic needs or immediate performance support is likely to result in the follower receiving substantial future goals (Deci et al., 2017; Hackman & Oldham, 1975; Herzberg, recognition, support, guidance, or rewards from the leader (Colquitt 1966). Moreover, these needs are particularly relevant in today’s et al., 2012; Lind, 2001; Van den Bos, 2001). As mentioned, since business environment characterized by competitiveness, rapid competence perceptions are more cognitive, they are potentially less technological changes, and flat hierarchies, where managerial focus directly pertinent to the development of these enduring affective is on employee knowledge, skills, and empowerment (Zhang et al., aspects of a high-quality hierarchical relationship (Moorman et al., 2012). Since autonomy-helping better satisfies these critical needs 2018). Thus, going beyond competence perceptions, we posit that of help recipients, we reason that providing it signals helpers’ unique leadership support can stem from the more affective and relationship- concern for recipients’ psychological needs and long-term well- relevant factor of perceived benevolence. being. That is, autonomy-helping, as compared to dependency- In support of this link, previous research indicates that perceived helping, signals greater levels of benevolence-based trustworthiness benevolence is conducive to high-quality relationships between on helpers’ part (hereafter, perceived benevolence). followers and current leaders (Dirks & Ferrin, 2002; McAllister, Overall, the above theorizing supports our prediction that though 1995; Yang et al., 2009). More relevant to our focus on supporting both help types signal greater benevolence than no help, autonomy- prospective leadership, cooperative group members who sacrificed (vs. dependency-) helpers typically convey to recipients that they can more of their own resources for the group were granted greater be trusted to be more benevolent, and as a result, recipients tend to influence and conferred higher social rank because they were seen as provide them greater support as their future leaders, as theorized next. motivated to promote group interests (Hardy & Van Vugt, 2006; Willer, 2009). Moreover, norm violators gained greater support for their leadership, but only when breaking rules to benefit others (Van Benevolence-Based Trustworthiness and Kleef et al., 2012). Together, the above findings are consistent with Leadership Support the social engagement hypothesis (Keltner et al., 2008) positing that people volitionally confer higher social ranks to those perceived as In volitionally supporting an informal leader, followers opt to caring more about, and advancing the interests of, their ingroup depend on another person who can influence their well-being and members. Applying this to the current context, and given that—as we attainment of needs (e.g., for mastery and autonomy, Bass & Avolio, theorized above—autonomy-helpers are typically perceived by 1993; Howell & Shamir, 2005; Kotter, 1990). That is, leadership recipients as even more benevolent than dependency-helpers, in support arguably carries considerable risk because it involves conjunction with the above findings and the social engagement questions of dependability and uncertainty about the potential hypothesis, we propose: leader’s future behaviors (Kramer, 2011; Lind, 2001; Van den Bos, 2001). It follows, then, that trust—a psychological state comprising the intention to accept vulnerability based on positive expectations Hypothesis 1: Autonomy-helping yields greater support for the of another person’s intentions or behavior (Rousseau et al., 1998)— helper’s leadership than does dependency-helping. in the potential leader is required to reduce this risk and enable follower support (Kramer, 2011). Although we know a good deal Hypothesis 2: The effect of help type on support for helper’s about the benefits of trust in current leaders (e.g., facilitating leadership is mediated by perceived benevolence. subordinates’ discretionary behaviors and job performance; Dirks & Ferrin, 2002; Kramer, 2011), the pivotal role of trust in the Next, pushing further our understanding of how recipients development of prospective leaders has been largely overlooked (cf. respond to the two help types, we examine a situational boundary Marinova et al., 2013). Addressing this lacuna, we highlight the condition for the advantage of autonomy-helping.  554 CHERNYAK-HAI ET AL. Boundary Condition: Time–Pressure Overview of Studies We have argued that autonomy-helping is typically more We tested these hypotheses in four studies conducted on advantageous for helpers’ perceived benevolence because it conveys employees placed in the role of help recipients. Employees working that they care for help recipients’ intrinsic needs and long-term well- at least 10 hr per week in a variety of sectors were recruited using being. Nonetheless, under conditions of urgency, there is strong various methods and sources: Israeli employees registered in an pressure for an immediate solution, increasing the importance of online panel (“Panel4All”; Studies 1 and 2b), snowball sampling via short-term performance goals relative to long-term autonomy and undergraduate students (Study 2a), and U.S. employees registered on mastery goals (Ballard et al., 2018; Beck & Schmidt, 2013). Thus, the Prolific platform (Study 3). Following best practices when using enhancing the attribution of benevolence to helpers providing online panel data (Porter et al., 2019), we provided information on dependency-helping under situational conditions that may justify it the panel data and reported both the time required to participate and (Jones & Davis, 1965), such as time pressure. the compensation (a minimum of $8 per hour). For Study 3, we also Specifically, based on attribution theory (Jones & Davis, 1965; restricted the sample to “high-reputation” workers (≥95% approval This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. Kelley, 1973), we theorize that when operating under time pressure rate). In all studies, participants were told that their purpose was to This document is copyrighted by the American Psychological Association or one of its allied publishers. (e.g., when completing a report under a tight deadline), help recipients explore employee behaviors and interactions in different workplace will attribute their helpers’ choice to offer dependency- (vs. autonomy-) situations. helping to this salient situational, rather than a dispositional, cause. All studies included instructional manipulation checks (Oppenheimer Further, when recipients are focused on the fear of missing a deadline et al., 2009), consisting of attentional checks incorporated within our or losing valued rewards due to slower performance, they may lack the key variables. Consistent with preregistrations, we excluded participants luxury, motivation, or available resources to learn or be empowered who failed instructional manipulation checks. In the simulation studies (Ballard et al., 2018; Beck & Schmidt, 2013). Thus, we posit that under (Studies 1 and 2b), we also excluded those who did not seek their time pressure, providing dependency-helping signals high care and teammate’s help or did not believe that the other teammate really sensitivity to the recipients’ short-term performance goals and thus existed. Finally, in Study 2b, we also screened out participants who did dependency-helpers will be perceived as highly benevolent. not care about receiving a bonus, rendering help less meaningful to Although this could suggest that sensitivity to salient performance them. Finally, in Study 3, we excluded employees who did not recall a needs under time pressure causes dependency- (vs. autonomy-) relevant helping event and thus did not undergo our experimental helping to signal greater benevolence (see Hypothesis 3b), we support manipulation (see Study 3, for details). The number of participants a different moderation prediction based on a competing argument. reported in each study is the number of valid participants after these Specifically, given our theorizing about the special care signaled by preregistered exclusions. autonomy-helping, we posit that autonomy-helpers will continue to Sample sizes (based on common practices for detecting mediation, be viewed as highly benevolent across situations. However, we Fritz & Mackinnon, 2007; Simmons et al., 2018; as well as on David further posit that dependency-helpers—typically perceived as less Kenny’s Shinyapp for mediation, with conservative estimates [.14] for benevolent—will under time pressure be viewed as more benevolent mediation paths,.80 power, and α =.05) were preregistered and strictly and as highly as autonomy-helpers. In sum, although we offer two followed. Given the preregistered exclusions and aiming to reach the competing hypotheses, we favor the prediction that the effect of help predetermined sample size, we targeted a slightly larger sample size. type (i.e., the advantage of autonomy-helping) will be eliminated See preregistration links under “Transparency and Openness.” under time pressure (Hypothesis 3a). We hypothesize a first-stage All studies used existing, validated scales to measure the variables. moderated–mediation effect (Figure 1): Unless noted otherwise, all items were rated on a 7-point scale (measures were preregistered). The results of a series of confirmatory factor analyses (conducted using Mplus, Version 8.4) supported the Hypothesis 3: The indirect effect of help type on support for discriminant validity of the key measures used in each study. helpers’ leadership via perceived benevolence is moderated by Specifically, the model in which each construct loaded on a separate time pressure, such that the advantage of autonomy- (vs. factor had a better fit than models combining constructs together or a dependency-) helping in regard to perceived benevolence (and one-factor model, Δχ2 ≥ 341.84 (Δdf ≥ 2), p <.01 (see Appendix A). thus, leadership support) is eliminated (Hypothesis 3a), or All studies manipulated help type (in a between-subjects design) reversed (Hypothesis 3b), for time-pressured tasks. but in different settings. Study 1, using a simulation, tested the main effect and mediation hypotheses (Hypotheses 1 and 2). Studies 2a and 2b, manipulating an additional between-subjects factor—time Figure 1 pressure—tested the full moderated–mediation model (Hypothesis 3; Conceptual Model see also Figure 1) using either a vivid role-playing workplace scenario in which the task deadline was either short or long (Study 2a) or a Time pressure virtual team simulation (similar to Study 1) involving (or not) the actual experience of acute pressure to provide a solution (Study 2b). Finally, Study 3, an experiment employing personal recollections Benevolence-based of workplace helping events (autonomy- vs. dependency-helping trustworthiness (Perceived benevolence) events), once again tested the main effect and mediation hypotheses to Autonomy- (vs. dependency-) Leadership provide greater external validity. support helping As theorized above, we focused on the role of perceived benevolence in mediating the effect of autonomy- (vs. dependency-) helping on  AUTONOMY-HELPING, DEPENDENCY-HELPING AND LEADERSHIP SUPPORT 555 leadership support. To further rule out alternative mediators and 1 addressed two additional, tangential issues. First, it was designed to demonstrate the robustness and uniqueness of this measure, we also confirm our assumption that both help types, compared to no help, accounted for competing mediators, measured after random assignment were more advantageous to helpers’ perceived benevolence and of participants into help-type conditions. Specifically, we controlled leadership support. Second, it examined whether receiving both help for perceived selfish motivation (Study 3) and perceived competence types together had an additional impact on perceived benevolence and (i.e., ability-based trustworthiness; in all studies) of the helper. leadership support beyond that from autonomy-helping alone (an Regarding the latter, though both help types clearly signal high exploratory analysis; see preregistration). Thus, Study 1 included competence, it may be that autonomy- (vs. dependency-) helpers are four conditions—dependency-helping, autonomy-helping, combined perceived as more competent or vice versa. Thus, we tested the helping, and no helping—in a between-subjects design. mediating role of perceived benevolence, beyond competence. We also controlled for recipients’ mood, measured in Study 1, Method which could reflect a simple positivity effect (e.g., Brief et al., 1995, for the impact of mood on employees’ job attitudes) in which Participants This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. participants view their helper as more benevolent and grant leadership This document is copyrighted by the American Psychological Association or one of its allied publishers. based on positive affect induced by a specific help type rather than by Participants were 5981 Israeli employees (338 women; Mage = perceptions of benevolence. Also, whereas for some tasks providing 40.12, SDage = 12.44) registered in Panel4All (an Israeli equivalent autonomy-helping is more time consuming or effortful (e.g., teaching to Prolific). Their mean organizational tenure (in months) was 61.78 a complex underlying process), for other tasks dependency-helping (SD = 104.00); their mean self-reported rank (from 1 = junior level may require more resources (e.g., completing an easy, but tedious, to 7 = senior level) was 4.08 (SD = 1.66); their mean weekly work task). In other cases, such as the decoding tasks used in Studies 1 and hours was 41.84 (SD = 13.29); and 78% reported having received 2b, providing either type of help may have similar time and effort higher education (>12 years). They worked in diverse sectors (21% requirements. Thus, we also measured participants’ perceptions of the science and technology, 7% education, 7% finance, 4% medicine, time and effort invested in the help (in all studies except 2a) and its and 7% administration). Participants were compensated with New perceived usefulness (Studies 2b and 3), to rule these out as alternative Israeli Shekel 7 (approximately $2) for completing the 10–15 min mediators and show unique mediation via perceived benevolence. simulation. For conciseness, in reporting our findings, we focus on our hypothesized mediator—that is, perceived benevolence. Additional Procedure and Measures robustness analyses also including the other helper’s perceptions— competence and selfishness—as competing mediators, as well as Participants were told that they would participate in a virtual team robustness tests supporting our mediator beyond all other alternative simulation consisting of two puzzle tasks, supposedly with another mediators (i.e., mood, perceived time, effort, and helpfulness), are participant—“teammate” (gender unknown)—connected online. To reported (per study) in Appendix C. increase psychological involvement and realism (Colquitt, 2008), they were told that correct and full answers would earn them an additional bonus beyond their basic participation reward (the bonus Transparency and Openness was delivered once data collection was completed). They were then We describe our samples, all data exclusions, all manipulations, told that in the first task, they were required to decode one sentence and all measures in each study, adhering to the Journal of Applied written in two secret codes (see Figure 2), thus offering high control Psychology’s methodological checklist. Study 1’s design, hypotheses, via the use of a single task and objectively manipulating help type— and analyses were preregistered at https://aspredicted.org/C24_JBW. tools (the two secret codes) versus full solution (the final decoding). Although Study 2a, using snowball sampling by undergraduate business The use of puzzles and anagrams is common in organizational administration students, was not preregistered, Study 2b, employing a research (e.g., Bonner et al., 2002; Locke & Latham, 1990; Porath & similar design, hypotheses, and analyses, was preregistered (see https:// Erez, 2007), and a decoding task was previously used to study aspredicted.org/NT1_7L6). Finally, Study 3’s design, hypotheses, and autonomy- versus dependency-helping in an organizational context analyses were preregistered (see https://aspredicted.org/XRX_MHH). (Montal-Rosenberg & Moran, 2022; Study 2). Participants were able All data, analysis syntax, and materials are available at https://osf.io/ to request their teammate’s help in decoding the sentence. We held gbd3v/?view_only=9b887f16e812438883b27b5be405c7c9. All studies the waiting time for teammate’s response constant (at 2 min) across were approved by Tel Aviv University’s ethics committee, Approval conditions. Number 0002105. Data were analyzed using IBM SPSS Statistics Next, participants were randomly assigned to a help-type (Version 27) and using Hayes’s (2017) PROCESS macro (Version 4.0), condition using a preprogrammed “teammate” response. Those in specifically Model 4 for mediation and Model 7 for moderated– the dependency- (vs. autonomy-) helping condition received the final mediation analyses. decoding (vs. the two secret codes, enabling them to decode the sentence independently). Those in the combined-helping condition received both the final decoding and the two secret codes. Participants Study 1 in the no-help condition received the message “No response has yet Study 1, a preregistered experiment, included a virtual-team-based 1 simulation in which participants received distinct types of help on a Among 628 participants who asked for their teammate’s help, consistent with our preregistration criteria, we excluded 19 who failed instructional uniform task. The study tested the effect of autonomy- (vs. manipulation check checks and 11 who did not believe that the other dependency-) helping on leadership support via perceived helper’s teammate really existed, resulting in a total of 598 valid participants (just benevolence (Hypotheses 1 and 2). Moreover, as preregistered, Study under our target sample size of 600 participants).  556 CHERNYAK-HAI ET AL. Figure 2 Manipulation Checks The Encrypted Puzzle Used in Study 1 We conducted a multivariate analysis to test the effect of help type on its two manipulation checks. The analysis revealed a significant effect, F(6, 1186) = 138.37, p <.001, Wilk’s λ =.35, η2p =.41, and help type affected both the autonomy- and dependency-manipulation checks, F(3, 594) = 161.86, p <.001, η2p =.45, F(3, 594) = 138.45, been received from your teammate,” whose wording was kept as p <.001, η2p =.41, respectively. Specifically, compared to participants neutral and as comparable to that of the other conditions as possible to in the dependency-helping condition, those in the autonomy- or avoid creating negative perceptions of the teammate or introducing combined-helping conditions indicated that they had received greater additional confounds (e.g., the suggestion that the helper purposefully autonomy-helping (B = 3.00, SE =.20, p <.001, η2p =.28, and B = withheld help). 2.60, SE =.20, p <.001, η2p =.22, respectively; Table 1). Compared to Then, we assessed our help type manipulation checks. Specifically, participants in autonomy-helping condition, those in the dependency- This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. we asked participants to describe the help provided to them using the and combined-helping conditions indicated receipt of greater This document is copyrighted by the American Psychological Association or one of its allied publishers. following items: help provided “tools for dealing with the current task dependency-helping (B = 1.05, SE =.22, p <.001, η2p =.04, and myself, ” “explanations on how to solve the current task myself” B = 1.48, SE =.22, p <.001, η2p =.07, respectively). Thus, our (Spearman–Brown =.89) for the autonomy-helping manipulation manipulations of the help type were successful. check, and “final solution to the current task,” “immediate solution to the current task” (Spearman–Brown =.91) for dependency-helping. Trustworthiness and Leadership Support We also measured participants’ perceived time and effort invested in As per our preregistration, to show that either type of help increased the help given to them, as well as their state positive affect (PA) and trustworthiness and leadership support compared to no helping, we state negative affect (NA), using the short Positive and Negative Affect conducted a multivariate analysis of the effect of help type on Schedule scale (Thompson, 2007; PA, α =.80; NA, α =.81), to rule out perceived ability and benevolence, as well as on leadership support. alternative mediators. All these measures were randomized in order. This revealed a significant effect of help type, F(9, 1431) = 35.89, p < As detailed below, participants were then asked to rate their.001, Wilk’s λ =.61, η2p =.15. Help type affected all three measures, teammate on perceived benevolence (our predicted mediator) and F(3, 590) = 122.00, p <.001, η2p =.38, F(3, 590) = 93.51, p <.001, ability (an alternative mediator); these measures were randomized in η2p =.32, and F(3, 590) = 82.54, p <.001, η2p =.30, respectively.3 As order. Finally, we told them that in the next task additional members expected, participants in the autonomy-, dependency-, and combined- would be included and a leader would be appointed for their team. helping conditions indicated greater trustworthiness and leadership We then measured their evaluations of the teammate’s suitability for support than those in the no-help condition, B ≥ 1.95, SE <.18, p < this leadership position..001, η2p >.17. Perceived Helper’s Benevolence (and Ability). To assess the two dimensions, we adapted items from Mayer and Davis (1999), changing the target of evaluation from “Top management” to “My Main Mediation Analysis teammate” and employing the Hebrew translations used in prior Following our preregistration, we used Hayes’s (2017) PROCESS research (SimanTov-Nachlieli et al., 2020). Furthermore, to better fit macro Model 4 to test the positive effect of autonomy- (vs. our one-shot simulation context, two irrelevant items were excluded dependency-) helping on leadership support (Hypothesis 1), via a priori: “Known to be successful at the things he/she tries to do” perceived benevolence (Hypothesis 2). In this analysis, help type from the Ability scale and “Would not knowingly do anything to hurt (coded using three dummy variables, with autonomy-helping as the me” from the Benevolence scale. Thus, as preregistered, participants reference group) was the predictor, leadership support was the rated the extent to which each of nine items applied to their dependent variable, and perceived benevolence was the mediator. teammate: four items gauged the teammate’s perceived benevolence Supporting Hypothesis 1, autonomy- (vs. dependency-) helping (e.g., “My teammate really looks out for what is important to me,” increased leadership support, B =.45, SE =.18, p <.05 (Table 2). α =.96), and five assessed perceived ability (e.g., “My teammate has Moreover, as predicted, it increased teammate’s perceived benevo- much knowledge about the work that needs done,” α =.97).2 lence, B =.39, SE =.19, p <.05, and perceived benevolence was Leadership Support. We adapted items from Stamkou et al.’s positively related to support for the teammate’s leadership, B =.67, (2019) scale (based on Rast et al., 2012) to measure leadership support in the next task, changing the target of evaluation (from “K” to “my 2 We focus here on differentiating between two primary trustworthiness helper”) to fit the current context. Furthermore, we excluded one item dimensions—ability and benevolence (Kramer, 2011)—which is consistent a priori (“I think K would represent the interests of the team very with well-established frameworks of social perception: competence and well”) due to its overlap with the mediator. Thus, as preregistered, six warmth (Fiske et al., 2007) or agency and communion (Abele & Wojciszke, items (rated from 1 = not at all to 7 = very much) assessed 2013). Notably, Mayer et al. (1995) included a third dimension—perceived integrity—which is the perception that the trustee adheres to a set of moral participants’ support for their helper’s leadership (e.g., “I would be a principles and is honest and truthful. This dimension is often included within strong supporter of my helper’s leadership,” α =.96). the communion (or warmth) dimension (along with benevolence) and seems less relevant here, given that in our context helping is instrumental and the helper is being truthful across conditions. Nonetheless, in a preliminary study Results and Discussion using personal recollections of helping events (see Supplemental Material), we found that the mediated effect of help type on leadership support via Means and standard deviations of all measured variables per perceived benevolence persisted when we controlled for this dimension. 3 experimental condition are provided in Table 1. Four participants did not complete the leadership support measure.  AUTONOMY-HELPING, DEPENDENCY-HELPING AND LEADERSHIP SUPPORT 557 Table 1 Means and Standard Deviations of All Measured Variables, per Condition (Study 1) Dependency-helping Autonomy-helping Combined helping No helping Variable M SD M SD M SD M SD Manipulation check—dependency 5.08a 2.02 4.05b 2.10 5.49a 2.04 1.47c 1.19 Manipulation check—autonomy 2.10a 1.49 5.10b 1.95 4.66d 2.09 1.57c 1.15 Perceived time invested in help 4.87a 2.09 5.35b 1.93 5.50b 1.98 2.30c 1.79 Perceived effort invested in help 4.77a 2.13 5.30b 1.82 5.29b 2.13 2.28c 1.73 PA 3.59a 0.79 3.46c 0.79 3.73a 0.75 3.44c 0.77 NA 1.69ac 0.79 1.53a 0.68 1.58ac 0.70 1.71c 0.76 Perceived ability 4.76a 1.59 5.06ab 1.75 5.36b 1.49 2.26c 1.38 Perceived benevolence 4.42a 1.61 4.81b 1.81 5.05b 1.61 2.23c 1.39 Leadership support 4.93a 1.59 5.38b 1.50 5.44b 1.54 2.98c 1.56 This article is intended solely for the personal use of the individual user and is not to be disseminated broadly. This document is copyrighted by the American Psychological Association or one of its allied publishers. Note. In each row, means with different subscript letters differ significantly from each other, p <.05. PANAS ratings (PA and NA) are on a 5-point scale. PA = state positive affect; NA = state negative affect; PANAS = Positive and Negative Affect Schedule. SE =.03, p <.001. Thus, supporting Hypothesis 2, the indirect rather, autonomy-helping is the crucial ingredient in signaling greater effect of autonomy- (vs. dependency-) helping on leadership support benevolence. via this mediator was significant, estimate =.26, SE =.13, 95% CI [.004,.529]. Studies 2a and 2b Notably, the effect of combined- (vs. autonomy-) helping on