|Home | About | Journals | Submit | Contact Us | Français|
The present research extends previous findings suggesting that sequential request techniques, such as the Foot-in-the-Door (FITD) or Door-in-the-Face (DITF) technique, are primarily effective under conditions conducive of mindlessness. We forward that this mindlessness may be the product of the influence technique itself. More specifically, based on the notion of self-control as a limited resource, we hypothesize that actively responding to the initial request-phase of a FITD-compliance gaining procedure drains the target of his/her self-regulatory resources, thus creating the mindlessness so often observed in social influence settings. This resource depletion opens the door for compliance with the target request. The results were in line with these expectations. More specifically, we observed that active responding to an initial request of a FITD technique reduced the availability of self-regulatory resources. This state of resource depletion mediated the effect of the technique on behavioral compliance. In addition, the results of this study ruled out the alternate explanation that the effects were attributable to mood or a general tendency for acquiescence.
Have you visited the city centre of your hometown lately to shop on a Saturday-afternoon? Then the following scenario has a good chance of accurately reflecting what has happened to you then. While chatting with your partner or friend, you spotted a group of representatives for a charity a few yards away and they have spotted you. Your first impulse was to avoid them by circumventing the group. Your attempt failed and before you could blink twice, you were already answering a few ‘harmless’ questions on what it meant to you to be a responsible citizen and which types of charity you already supported. Then the target request for a (substantial) financial donation to a novel cause was presented. Have you complied? Then you are not alone! Persuasion agents such as fundraisers, sales representatives, advertisers, politicians, and marketers are often stunningly successful in feeding the flame of desire or reducing resistance to influence on the part of the target consumer. A host of social influence techniques can be employed to get people to say “yes” to an offer or request they were not planning to yield to in advance (Cialdini and Trost 1998), and more often than not, these techniques accomplish what they were set out to do. But what is the key mechanism responsible for this effectiveness? And why is it that we sometimes, or even in most instances, do not consciously realize that a sales trick is being played on us, and we are falling for it, head-first? A key objective of the present research is to address these questions.
During the past four decades, a variety of influence techniques have been studied, including the Foot-In-The-Door technique (Freedman and Fraser 1966), the Door-In-The-Face technique (Cialdini et al. 1975), and—more recently—the Disrupt-Then-Reframe technique (Fennis et al. 2004). In recent years, this research has increasingly emphasized processes that are subtle, indirect and outside conscious awareness of the target. More specifically, the notion of automaticity or “mindlessness” (Langer 1992) has been forwarded as the cornerstone of all influence (Cialdini and Goldstein 2004). Under these conditions of reduced mental alertness, the target is thought to fall back on habit and routine and hence will employ ‘shortcuts’ or simple heuristics to arrive at a decision. Use of these heuristics will generally increase the likelihood of compliance (Cialdini 1993). Heuristics the target will resort to in these circumstances include the principles of consistency (the propensity to show congruent behavior across situations), reciprocity (the felt obligation to return a favor) or liking.
Indeed, research is replete with examples of the use of these heuristics in social influence situations (see Cialdini and Goldstein 2004 for an overview). For instance, several studies on the Foot-In-The-Door (FITD) technique (whereby the target is initially presented with a small request, followed by a more substantial request) have shown that employing the technique generally results in increased compliance, primarily because the target wants to behave consistently across situations (see Burger 1999). That is, compliance with the first, small request, induces the self-perception that one “is the kind of person to comply with these kinds of requests” which increases the odds of compliance with the more substantial second request. Similarly, Cialdini et al. (1995) found that individuals scoring high on the Preference for Consistency-scale were more likely to show compliance with an FITD-type of request than individuals low in this preference. The need to behave consistently across conditions, regardless of careful scrutiny of the exact nature of the various requests, thus constitutes an important heuristic principle that people use when presented with influence techniques, but primarily when they are in a state of mindlessness. Indeed, research on the related “That’s-Not-All” technique (Pollock et al. 1998), whereby an initial request is followed by a series of subsequent requests that are gradually made more desirable to the target, has shown that its effectiveness is reduced when the target’s mental alertness increases and he or she becomes “mindful” (Langer 1992).
Another powerful heuristic that people draw upon when in a state of mindlessness is the motivation to return a favor, i.e. the principle of reciprocity (Gouldner 1960). A technique that hinges on this principle is the Door-In-The-Face (DITF) technique which is characterized by a sequence of rejection-then-moderation (see O’Keefe and Hale 2001). In the DITF-technique, a large request (which probably will be rejected) is followed by the milder target request. Although not uncontested, it is generally assumed the technique works because the influence agent, by sizing down the request, makes a clear concession. For the target, this would evoke the need to make a concession in return and thus to comply with the milder request. In a classic study, Cialdini et al. (1975) provided a first demonstration of the effectiveness of the DITF technique. They asked Arizona State University students to work as non-paid volunteers for the Juvenile Detention Center. Everyone rejected. Next, however, the request was made milder: they were asked to take a group of juveniles to the zoo for 2 h. Under these conditions, half of the participants complied with the request. In the target request-only control condition, where participants were only asked to take the juveniles to the zoo for 2 h, only 17% complied. A follow-up study more directly tested the reciprocity account. In that study, participants were asked by either one or two influence agents to volunteer to assist low-income children for two years. After rejection, participants were asked to take the group of children to the zoo for 2 h. Compared to control conditions, the DITF condition again proved effective in bringing about compliance, but only when the same individual made both requests. When one person made the large request and the second person the smaller request, the technique ceased to be effective. In the case of a single requester scenario, the agent made a clear concession, which evoked the norm of reciprocity, whereas in the two-requester scenario this norm was not made salient. Importantly, this norm operates under mindless conditions. Hence, the DITF technique, too, is more effective when participants are mindless, rather than mindful.
In sum, it is evident that the effectiveness of a broad variety of social influence techniques hinges on the principle of mindlessness. But where does this mindlessness in these and other social influence situations stem from? In other words: what is it that makes people behave ‘automatically’ and fall back on engrained heuristics when presented with an influence technique under these circumstances? Although mindlessness has been proposed as a “conditio sine qua non” for the techniques to work, it is interesting to note that research to date has failed to directly address this key question.
We argue that the origins of the concept of mindlessness, and thus of the effectiveness of many social influence techniques, frequently reside in the influence setting itself, rather than in extraneous factors. More specifically, we forward that the origins of mindlessness can be found in a characteristic that most techniques have in common: multiple decision moments, or sequential requests. That is, the target of influence has to make one or several decisions, or choices, before the target request is presented. For example, the Foot-In-The Door procedure starts with a small request, followed by a larger request. Similarly, the Door-In-The-Face technique starts with a relatively large request, followed by a smaller request. We propose that all sequential request techniques essentially trigger one underlying psychological mechanism that accounts for their impact: that of self-regulatory resource depletion (Baumeister et al. 1998; Fennis et al. 2009; Muraven et al. 1998; Vohs and Heatherton 2000). The basic premise of resource depletion (also termed ego-depletion) is that self-control processes such as actively responding to influence attempts, exercising self-control, or using will power require resources that are finite: hence, the active self can become depleted. Given that the initial request-phase of sequential request techniques often involves active responding, such as answering several questions, agreeing to receive and process a persuasive message, or, conversely, resisting or rejecting an opening offer, (Fennis et al. 2004), we propose that the target’s self-control resources will be taxed in the opening stages of the influence attempt. Hence, the process of actively responding to the initial request draws upon this same limited resource. We view responding to social influence as a manifestation of self-regulation requiring resources, and compliance as self-regulatory failure brought about by resource depletion. The following section reviews research germane to our notions.
Similar to the functioning of a muscle, the resource depletion model (Baumeister et al. 1998) posits that any form of behavior that involves deliberate, conscious and controlled responses by the self, draws on a limited resource, akin to strength or energy. In addition, one act of volition is posited to have a detrimental impact on any subsequent act that draws from the same resource. As a consequence, and similar to muscle failure after repeated straining, a series of self-regulatory acts will deplete the resource up to the point of self-regulatory failure (Baumeister 2002). This resource can be replenished but at a slower rate than it is consumed. Hence, recuperation is possible, but this will take some time. In a condition of self-regulatory failure, the self is less able to function effectively which results in reliance on habit, routine, and automatic processes, key indicators of mindlessness (Baumeister et al. 2000).
Importantly, Baumeister et al. (1998) have stated that making choices and decisions directly draws on this same resource (see also Vohs et al. 2008). This notion was examined in an experiment where participants were requested to deliver a counter- or pro-attitudinal speech, either as an act of free choice (high volition) or forced to do so (low volition; Baumeister et al. 1998, exp. 2). The key dependent measure to assess the extent of depletion consisted of a problem-solving task involving a difficult puzzle. In support of predictions, it was found that freely complying with the request to deliver the speech (regardless of attitudinal position) resulted in lower persistence and fewer attempts in solving the subsequent puzzle than when participants were forced to deliver the speech. In a more recent series of studies, Schmeichel et al. (2003, exp 1) have demonstrated that devoting attention to a demanding task also robs the self from its resources. In their study, Schmeichel et al. (2003) showed that participants who were asked to actively concentrate on a videotape performed worse on subsequent (unrelated) measures of cognitive functioning than control participants. Importantly, these manifestations of reduced critical scrutiny and analytical reasoning are also endemic to conditions of mindlessness (Langer 1992). Interestingly, a recent meta-analysis of more than 30 years of research on the FITD technique (Burger 1999) has revealed that the effectiveness of the technique is dependent on the level of involvement with the initial request. Fully compatible with our depletion account, the FITD appeared more effective when the initial request is more cognitively demanding and requires more active (initial) decision-making. For instance, Tybout (1978, Experiment 1) asked participants to simply sign a petition, or asked them to explain to the influence agent their personal reasons of why they signed (high involvement). Pliner et al. (1974) examined compliance with a donation request and preceded the target request either by an initial request asking participants to wear a daffodil pin or asking them to wear the pin and also persuade family members to wear the pin. Presumably, the act of actively persuading others to wear the pin requires self-regulation since the target is required to muster arguments to convince others and present him/herself in a favorable and socially desirable light. In addition, Fish and Kaplan (1974) asked participants either to listen to a lecture (low demanding initial request) or craft and write an essay (high demanding initial request) before the target request was posed. Moreover, Seligman et al. (1976) asked for responses to five initial questions regarding “people’s reaction to the energy crisis” versus 20, 30, or 45 questions (responding to more questions constituted a more demanding task). In these studies, compliance with the target request was higher when the initial request demanded more intellectual processing as opposed to when it was less intellectually demanding. Finally, the pioneering work on the FITD technique by Freedman and Fraser (1966) also shows that more involving or demanding initial requests produce more compliance with the target request. In their first foot-in-the-door experiment, Freedman and Fraser (1966) approached households and before the larger target request was posed (i.e., a request to volunteer as a research participant in a large survey on household products), participants were asked whether they agreed to answer eight questions about the kinds of soaps they used. Importantly, the authors also varied the extent of performance required with respect to the initial request. That is, participants either proceeded to actually answer the initial questions or only agreed to do so. In line with the previous studies, the results showed that compliance with the target request was higher when participants had actually answered the series of questions rather than only agreeing to do so.
From a resource depletion perspective these findings makes good sense: a more demanding and absorbing initial request will deplete resources to a higher extent than a less salient, inconsequential one, thus lowering resistance to the subsequent target request.
In sum, there is accumulating evidence suggesting that actively responding to the initial request phase of a FITD technique may affect the extent of resource depletion. This state of resource depletion may capture the mindlessness endemic to many social influence contexts. In addition, resource depletion has been demonstrated to facilitate less deliberative, more impulsive and heuristic decision making (see Fennis et al. 2009). Translated to the present context, we therefore expect that a more cognitively demanding FITD script will produce higher compliance rates than a less demanding script. Moreover, we hypothesize that the type of FITD script will affect the extent of target resource depletion. Finally, we expect self-regulatory resource depletion to mediate the impact of the technique on compliance.
We tested our hypotheses in a single factor design with type of influence technique (demanding vs. undemanding initial request) as between-subjects factor and self-regulatory resource depletion and compliance as dependent variables. A total of 56 individuals (47 male and 9 female) with a mean age of 21.4 years (SD=2.86) participated in the experiment.
Participants were approached by three confederates (two males, one female) claiming to be members of a student action group that aims to improve academic education. They asked participants to take part in a short study on the issue conducted by the group in return for 2.50 Euros. After they agreed, participants were seated in laptop-equipped cubicles where the entire experiment took place, and all instructions were presented.
We used a subtype of the FITD compliance procedure termed the ‘continued questions procedure’ (CQP). In this procedure, the target request is preceded by a smaller initial request. This initial request consists of a number of probing questions which are semantically related to the target request (see Burger 1999). Hence, both the initial request and the target request pertained to the issue of improving education. Previous research has shown that this type of FITD technique is effective in promoting compliance (see Burger 1999; Fennis et al. 2009). The technique hinges on the heuristic principle of consistency and commitment (Cialdini 1993). The target is thought to infer a positive attitude from his/her active responses to the initial questions which promotes a favorable disposition to agree with the target request (Burger 1999). Note that our reasoning implies that such heuristic decision-making is particularly likely when the target is deprived of his/her self-control resources and consequently is in a state of mindlessness. Moreover, we expect and test that it is the act of actively responding to the initial questions that drains the self of these resources.
Typically, to assess its impact, the CQP is contrasted with a control condition where the initial request stage (i.e., the initial set of probing questions) is omitted (Burger 1999). However, in the present context this could inadvertedly introduce a design confound, because the number of times the target is required to respond and the interaction duration may then covary with the cognitive effort involved in responding. Hence, in the present study, we kept the basic structure of the CQP constant across conditions and only manipulated the extent to which answering the initial questions would drain self-regulatory resources. In addition, as a confound check we also measured our participants’ mood.
After a general introduction of the student action group and its objectives, participants were randomly assigned to the demanding or undemanding initial request of the CQP procedure. In all conditions, the participant was presented with an initial request asking them to rate their agreement with a series of four statements on academic education. These statements argued for an increase in the number of weekly hours that students should have class, the requirement for students to pay back government funded loans and a fine in case of failing to meet the graduation criteria in four years, a proposal to cut down the number of fixed computers on campus and instead offering students the possibility of buying a personal laptop at reduced rate, and the proposal to increase the basic student grant awarded by the government to meet students’ financial demands.
Participants in the demanding initial request condition were then asked to generate three arguments opposing their position on each of the statements. For example, when participants indicated to agree with the statement to raise the number of weekly hours devoted to classes, they were asked to come up with three reasons why they should disagree with the statement (and vice versa). Hence, all participants in this condition were asked to actively override their primary evaluative response to the statement, an act requiring active self-regulation (Wheeler et al. 2007). In contrast, participants in the undemanding initial request condition were simply asked to generate three arguments supporting their position on each of the statements. There was no time limit for answering. A MANOVA showed that the time required to answer each of the four statements was not influenced by type of influence technique (F<1).
Self-regulatory Resource Depletion We used the Stroop-task to assess self-regulatory resource depletion (Stroop 1935; Vohs et al. 2005; Wallace and Baumeister 2002). In the Stroop task, participants are exposed to color words that are depicted in either a corresponding or a deviating font color and they are asked to report the font color. Earlier research on self-control has found that the Stroop task taxes self-regulatory resources (Gailliot et al. 2007; Muraven et al. 2006; Webb and Sheeran 2003; Wallace and Baumeister 2002). Hence, for participants already low in self-regulatory resources it is harder to inhibit or override their incipient response to read the word and instead to report the font color. As a result, depleted participants will show impaired Stroop performance and thus make more errors and report less correct answers than their non-depleted counterparts.In line with previous research (Fennis et al. 2009), participants responded to each stimulus by clicking one of four buttons on their computer screen, which corresponded to the various color words. They received 32 randomized trials, of which 8 were congruent (a stimulus word was presented in a font color that matched its semantic meaning; e.g., “blue” was presented in blue font) and 24 that were incongruent (a stimulus word was presented in a font color that mismatched its semantic meaning; e.g., “blue” presented in red font). Participants were instructed to report the font color of each word as quickly and accurately as possible. The number of correct responses was recorded and served as a measure of self-regulatory resource depletion (cf. Gailliot et al. 2007).
Mood To measure whether the type of influence technique would result in unintended mood effects (e.g., overriding one’s initial position on the statements may evoke negative affect), we administered the Positive and Negative Affect Schedule (PANAS; Watson et al. 1988), which includes 10 positive and 10 negative affect-items using a 5-point scale. An index of positive and negative mood was created by averaging the scores on the respective items (α=.82 for the positive affect index and α=.72 for the negative affect index).
Compliance At the ostensible end of the study, participants were thanked for their collaboration. Then the target request was posed. In line with the requirements for a CQP, this target request solicited compliance on an issue related to the initial request. Participants were asked to volunteer in future events and endeavours by the student action group. More specifically, the target request read: “to further improve the quality of education, the student action group “Better Education” is looking for volunteers to attain its goals. If we would ask you, how many hours are you willing to volunteer this year?” Participants could indicate the number of hours between 0 and 5 that they were willing to volunteer (M=1.20, SD=.52). Number of hours served as our measure of compliance.
A first series of two ANOVA’s was performed to ascertain that our manipulation of type of influence technique did not unduly influence participants’ positive or negative mood states. Results of this analysis ruled out differential affect as an alternative explanation, since the impact of influence type failed to reach significance, both on the index of positive affect (F (1, 54)=1.71, n.s.) and on the index of negative affect (F<1).
In line with our hypothesis and previous research on the FITD, the type of influence technique affected compliance rates (F (1, 54)=5.83, p=.02). Inspection of the means showed that participants responding to a more cognitively demanding initial request were willing to donate more time to the student action group (M=1.36, SD=.68) than participants who responded to the less demanding initial request (M=1.04, SD=.19).
In addition to the effect of type of influence technique on compliance, analysis of variance revealed a main effect on the extent of resource depletion as indicated by Stroop performance (F (1, 54)=4.88, p=.03). This effect showed that responding to a more demanding initial request produced higher levels of resource depletion than responding to an undemanding initial request. More specifically, the demanding initial request resulted in less correct answers on the Stroop task (M=29.04, SD=6.30) than the undemanding initial request (M=31.68, SD=.67).
Finally, to determine whether the extent of resource depletion indeed mediated the impact of the FITD technique on compliance, we performed a mediation analysis (cf. Baron and Kenny 1986). The results of the analysis of variance reported earlier had established that the type of influence technique (the independent variable) affected compliance rates (the dependent variable) and the extent of resource depletion (the proposed mediator). A final regression analysis with compliance rates as criterion and type of influence technique (dummy coded) and extent of resource depletion (centered, see Aiken and West 1991) as predictors showed that the impact of resource depletion was significant (β=−.29, t(55)=−2.21, p<.03) whereas the previously significant effect of type of influence technique on compliance was reduced to non-significance (β=.23, t(55)=1.76, n.s.), thus indicating full mediation (Baron and Kenny 1986).
The present results demonstrate that the roots of mindlessness endemic to social influence contexts, can be a property of the influence setting itself, rather than the product of some extraneous factor. Earlier findings had already yielded support for the notion that mindlessness is an essential condition for influence scripts to “work” and affect consumer compliance (Cialdini 1993). We showed that it is the act of actively responding to the initial request phase of a sequential request technique that creates this state of increased susceptibility to influence. More specifically, the present results corroborate the notion that once a target of influence agrees to be exposed to the initial request phase, he/she opens the door to succumbing to the ultimate target request. Our results suggest that social influence strategies that are comprised of a series of requests all work to trigger one underlying process: that of self-regulatory resource depletion. That is, actively responding to the initial request (be that answering a series of cognitively demanding questions as in the present case, or actively rejecting an opening offer as in the DITF-technique) drains the target of its self-regulatory resources. A reduced supply of regulatory resources, in turn, fosters compliance with the actual target request through reliance on salient heuristics. In short, responding in an effortful way to an initial request induces self-regulatory resource depletion, which subsequently encourages heuristic decision-making. In the present research this heuristic decision-making was the result of the principle of commitment/consistency which is embedded in the CQP and other types of FITD scripts.
Our findings show that self-control resources are a pivotal intervening construct in the process of compliance. As such the present results nicely align with recent studies in the neighbouring field of attitude change in response to persuasive communication, which have shown that resisting persuasion similarly requires self-regulation and hence becomes more difficult when message recipients are in a state of depletion-induced mindlessness (Burkley 2008; Wheeler et al. 2007). In addition, our findings extend earlier research on the role of self-control in social influence (Fennis et al. 2009; Janssen et al. 2008) in demonstrating the pervasive role of self-regulatory resources in a new domain of compliance gaining. More in particular, whereas these previous studies focused on charitable and prosocial behavior, the present influence setting pertained to political activism, albeit on a small scale (i.e., a student action group arguing for educational reforms and soliciting volunteers). Our findings imply that such activism may not always be fueled by strong political convictions or motivations, but is also spurred when people are deprived of their self-control resources and thus become vulnerable to the activist’s influence attempt targeted at them. In a more dramatic context, this process has also been suggested by Baron (2000) in his influential treatise on coercive political persuasion and indoctrination. More specifically, Baron (2000) pointed out that such coercive programs typically start with a “softening up”-stage in which depletion-inducing factors such as disorientation, stress, or sleep deprivation break down initial resistance by the target. As such, the role of self-regulation and self-regulation failure may constitute a viable and novel perspective to shed light on the dynamics involved in (group) activism, radicalization, religious cults, and possibly even terrorism.
The present results showed that resource depletion fully mediated in the impact of the FITD technique on compliance, suggesting that self-control is a necessary and sufficient condition for heuristic decision making in social influence settings. Moreover, our research effectively ruled out (negative and positive) mood as an alternate explanation. Finally, the results suggest that compliance in these sequential request settings cannot be attributed to a general tendency for acquiescence. Indeed, the CQP procedure used in our research produced the highest levels of compliance when participants were asked to argue against their positions on the initial statements rather than in support of them. Hence, the commitment/consistency principle triggered by the CQP technique appears to pertain to the effort devoted to the initial response, rather than the valence of this initial response per se. That is, the target infers from his/her initial level of expended effort that he/she seems to care about the issues advocated by the activist group and hence is prepared to invest more time and effort as a volunteer for the group. Future research might profitably explore this possibility more in depth, for instance by systematically varying the valence of the initial and target request of a sequential request technique in addition to the involvement with the initial request.
Open Access This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.
Bob M. Fennis, Phone: +31-50-3637065, Fax: +31-50-3638252, Email: firstname.lastname@example.org.
Loes Janssen, Phone: +31-13-4662154, Fax: +31-13-4662067, Email: email@example.com.