The missing link? Implications of internal, external, and relational attribution combinations for leader–member exchange, relationship work, self‐work, and conflict

• Published date: 01 June 2019
• Date: 01 June 2019
• DOI: http://doi.org/10.1002/job.2349

SPECIAL ISSUE ARTICLE

The missing link? Implications of internal, external, and

relational attribution combinations for leader–member

exchange, relationship work, self‐work, and conflict

William L. Gardner |Elizabeth P. Karam |Lori L. Tribble |Claudia C. Cogliser

Rawls College of Business, Texas Tech

University, Lubbock, Texas

Correspondence

William L. Gardner, Rawls College of Business,

Texas Tech University, 703 Flint Avenue,

Lubbock, TX 79409.

Email: william.gardner@ttu.edu

Summary

Attributions are causal explanations made by individuals in response to important,

novel, and/or unexpected events. Numerous attribution theories have examined

how people use information to make attributions and how attributions impact an

individual's subsequent emotions and outcomes. However, this research has only

recently considered the implications of dyadic‐level attributions (i.e., relational

attributions), particularly in the context of leader–follower relationships in organiza-

tions. Therefore, the purpose of this theoretical paper is threefold. First, we integrate

research on attributional biases into the research on relational attributions. Second,

we integrate and extend attribution theory to consider the implications of convergent

and divergent internal, external‐person, external‐situational, and relational attribu-

tions for leader–member exchange (LMX) quality, relationship work, self‐work, and

conflict. Third, we make the implicit ranking of attribution combinations and the resul-

tant levels of relationship work explicit. In doing so, we contribute to attribution the-

ory and research by proposing how attribution combinations produce positive and

negative outcomes that are both intrapersonal and interpersonal. Further, we contrib-

ute to the LMX literature by explicating how leader–follower attribution combinations

influence relationship quality.

KEYWORDS

attribution theory, conflict, leader–member exchange (LMX), relational attributions, relationship

work

1|INTRODUCTION

“Naive psychology”; Heider (1958), the founder of attribution theory,

used this term to convey his assertion that humans have an innate

desire to attribute causes to events (p. 5). When making such

attributions, Heider posited that people make a fundamental distinc-

tion between internal and external causes for events that occur.

By definition, internal attributions (e.g., ability and effort) arise from

the focal actor, whereas external attributions emanate from the

situation (e.g., luck and task difficulty; external‐situational attributions)

or other parties (external‐person attributions). Throughout the rich

history of attribution theory (e.g., Martinko, 1995, 2004), these two

poles of the locus of causality dimension—internal and external—have

remained at its core.

Recently, Eberly, Holley, Johnson, and Mitchell (2011) proposed a

refinement to attribution theory by introducing a third category of

locus of causality attributions that pertains to the dyadic level of anal-

ysis: relational attributions. Specifically, they argued that people in

relationships sometimes attribute the cause of an event to the rela-

tionship with the other person (i.e., “something about us”; a relational

attribution) rather than to purely internal or external sources. They

defined relational attributions as “those explanations made by a focal

individual that locate the cause of an event within the relationship that

the individual has with another person”(p. 736). They also suggested

Received: 5 January 2018 Revised: 18 December 2018 Accepted: 23 December 2018

DOI: 10.1002/job.2349

554 © 2018 John Wiley & Sons, Ltd. J Organ Behav. 2019;40:554–569.wileyonlinelibrary.com/journal/job

that when relational attributions are made, the parties are more likely

to engage in relationship‐focused behaviors, or relationship work.In

subsequent empirical research using data from seven samples, the

authors tested and found support for the existence of relational attri-

butions that are distinct from internal and external attributions, and

they found that relational attributions are related to relationship work

(Eberly, Holley, Johnson, & Mitchell, 2017).

In considering the implications of relational attributions, Eberly

et al. (2011) presented arguments for how attributions can prompt

individuals to make personal changes (i.e., enhance skills), situational

changes (i.e., change jobs), or relationship changes (relationship work).

These adaptive behaviors are suggested to result from internal,

external, and relational attributions, respectively. The theoretical

arguments, however, have yet to consider how an individual's

attributional biases impact the type of attribution that is made. The

exclusion of biases from relational attribution theorizing is noteworthy

because biases have the potential to significantly impact an individ-

ual's perception of an event. Moreover, “biases in the attribution

process may be an important source of perceptual conflict between

leaders and members”(Martinko & Gardner, 1987, p. 239). Therefore,

our first purpose is to integrate attributional bias and relational

attribution theory.

In addition, relational attribution theory has not explored how

leader–follower attribution combinations may have implications for

relationship work as well as other important outcomes. The relation-

ship between leaders and followers is often one of the strongest

connections between employees (Hui, Lee, & Rousseau, 2004). These

relationships are commonly characterized by a high degree of interac-

tion and task dependence along with expansive time bracketing—a

history of past interactions and an expectation of future exchanges

(Colquitt et al., 2013). Therefore, unexpected and negative events

are likely to prompt a search for causes by both the leader and

follower in this important organizational relationship, and in response

to the resultant attributions, emotions and behaviors of both parties

are likely to be provoked. Attribution theory has long established the

link between emotion and behavior (Weiner, 1985), and prior

theoretical research has explored leader–follower attribution

responses (e.g., Martinko & Gardner, 1987). However, what is missing

is a comprehensive consideration of the responses to leader–follower

attribution combinations that may be either convergent (i.e., the leader

and follower make the same attribution for the event) or divergent (i.e.,

the leader and follower make different attributions for the event).

Therefore, our second purpose is to integrate and extend the theory

by Martinko and Gardner (1987) and Eberly et al. (2011) to consider

the implications of convergent and divergent internal, external‐person,

external‐situational, and relational attributions for leader–member

exchange (LMX) quality, relationship work, self‐work, and conflict. In

addition, we make the implicit ranking of attribution combinations

and the resultant levels of relationship work explicit. We believe that

examining these attribution combinations may provide the “missing

link”for better understanding both intrapersonal and interpersonal

leader–follower outcomes.

Thus, the manuscript proceeds as follows. First, we provide an

overview of attribution theories and LMX theory. Next, we present a

model of leader and follower attributional biases, attribution

combinations, LMX quality, relationship work, self‐work, and conflict.

We proceed by integrating the work of Martinko and Gardner

(1987) and Eberly et al. (2011) to discuss the implications of leader

and follower attribution combinations for these outcomes. Then, we

discuss the interaction of attribution combinations and LMX quality

to describe how LMX quality impacts the likelihood of relationship

work and conflict. Finally, we discuss theoretical and practical implica-

tions of the propositions as well as areas for future research.

2|THEORETICAL FOUNDATIONS

2.1 |Attribution theories

“Attributions are the causal explanations that individuals use to inter-

pret the world around them and adapt to their environment,”espe-

cially when attempting to understand events that are important,

novel, unexpected, and negative (Eberly et al., 2011, p. 733). Scholars

have remarked that “there is no ‘attribution theory’…[rather] the term

attribution theories applies to a broad group of theories that are con-

cerned with causal reasoning”(Martinko & Thomson, 1998, p. 272).

Attribution theories, therefore, focus on the processes through which

people attempt to explain the causes of behaviors or events (Harvey,

Madison, Martinko, Crook, & Crook, 2014; Heider, 1958; Kelley,

1967, 1973; Martinko, Harvey, & Dasborough, 2011; Weiner, 1985).

Numerous scholars have extended Heider's (1958) initial work on

attributions. Below, we describe several attribution theories that are

most relevant to our model.

Kelley (1967, 1973) introduced the covariation model to describe

how people make attributions using informational cues across multiple

observations. He suggested that individuals consider three informa-

tion criteria—distinctiveness, consistency, and consensus—related to

a behavior or event when making attributions. Distinctiveness infor-

mation involves within‐person behavior that “compares the behaviors

of the individual in other situations”(Martinko & Thomson, 1998, p.

273). Consistency information also involves within‐person informa-

tion, but it incorporates a longitudinal perspective that considers

whether the behavior is similar or different across time (Kelley,

1973; Martinko & Thomson, 1998). Consensus information assesses

whether behavior is common across individuals in similar situations;

that is, it is a between‐person appraisal that compares the focal party's

behavior with the behavior of others (Harvey et al., 2014; Kelley,

1973). When consensus is high, it generally leads to an internal locus

of causality, whereas low consensus information leads to an external

locus of causality. When combined, Kelley's three informational

criteria form patterns that individuals use to attribute causes to spe-

cific behaviors or events. Ability attributions generally occur when

there is information that is high on distinctiveness, high in consistency,

and low on consensus. Attributions related to a person's effort are

likely when information is low on consensus, low on consistency,

and high on distinctiveness. High consensus information leads to an

external locus of causality. Information about the target that is high

in consensus, consistency, and distinctiveness will likely result in

attributing the outcome to the nature if the task. An outcome that is

high on consensus and low in both consistency and distinctiveness will

GARDNER ET AL.555