Inducing Compliance With a Request: the List Technique

Peter H. Reingen, University of South Carolina
ABSTRACT - A compliance induction technique where a target is first shown a list of other compliers and then is asked to comply with a request was examined in a field experiment. The technique was found to be effective provided that the number of other compliers was sufficiently large. A formulation based on informational social influence is suggested to account for the findings.
[ to cite ]:
Peter H. Reingen (1979) ,"Inducing Compliance With a Request: the List Technique", in NA - Advances in Consumer Research Volume 06, eds. William L. Wilkie, Ann Abor, MI : Association for Consumer Research, Pages: 45-49.

Advances in Consumer Research Volume 6, 1979      Pages 45-49


Peter H. Reingen, University of South Carolina

[The author would like to thank the Heart Association for its support. This paper reports a portion of a larger study.]


A compliance induction technique where a target is first shown a list of other compliers and then is asked to comply with a request was examined in a field experiment. The technique was found to be effective provided that the number of other compliers was sufficiently large. A formulation based on informational social influence is suggested to account for the findings.


A growing number of investigations have recently focused on variables that affect a person's willingness to comply with a request. Studies have been conducted to examine the effects on compliance of such factors as prior compliance with a small request (e.g., Cann, Sherman, and Elkes 1975; Freedman and Fraser 1966; Reingen in press; Reingen and Kernan 1977; Scott 1977); prior noncompliance with an extreme request (Cialdini, Vincent, Lewis, Catalan, Wheeler, and Darby 1975; Cialdini and Ascani 1976; Reingen in press); cost (e.g., Wagner and Wheeler 1969); need (e.g., Wagner and Wheeler 1969); incentives (e.g., Scott 1976); commitment then cost (Cialdini, Cacioppo, Basset, and Miller in press); legitimization of paltry contributions (Cialdini and Schroeder 1976; Reingen 1977; Reingen in press); and observation of a complier (e.g., Bryan and Test 1967; Wagner and Wheeler 1969).

The last of these factors, observation of someone complying with a request (i.e., a model), seems at first glance the least promising from a pragmatic viewpoint. To enhance the probability of a target's yielding to a request via his observation of a model's behavior would indeed appear to be a rather cumbersome and probably inefficient technique in many large-scale request contexts of interest to consumer behaviorists, such as fundraising drives, blood drives, political campaigns, brand marketing, and the like. Yet, there are techniques employed in everyday compliance settings that bear more than a superficial resemblance to modeling procedures: when a textbook publisher provides professors with a list of new book adopters, when names of contributors to a charity drive are announced during a telethon, and when people are asked to add their name to a petition, to mention but just a few.

The common ground between the everyday practices and the modeling procedures in social psychology appears to be that both are presumably predicated on the efficacy of informational social influence; namely, that a target often uses the behavior of another (or others) to assist him in determining his own action (Deutsch and Gerard 1955). What primarily distinguishes the two is that only the cognate modeling research in social psychology involves a target's observation of a present other's compliance with a request; i.e., the modeling effect appears to be mediated by direct informational social influence. In the case of the common practices, a requester merely provides cues to a target that others have complied; i.e., a compliance effect, if obtained, may be mediated only indirectly by informational social influence.

But does a technique where a target is merely informed of other compliers by a requester really work, or have practitioners, in the absence of hard evidence, deluded themselves as to the compliance-producing power of the technique? If it does work, is informational social influence the underlying mediator of the effect?

To provide answers to these questions, a study was conducted in a naturalistic setting. Targets (college students) were asked to donate money to the Heart Association in one of ten ways. In the first (the control) a direct request for a donation was made. In the next four conditions subjects were first shown a list of (fictitious) other contributors and then were asked for a donation. Two factors were manipulated: Size-of-donations from others and number-of-donors.


It has been well documented that information dependency on others tends to increase with uncertainty (King 1975). Since students are not the typical target of fund-raising efforts, their experienced uncertainty with regard to how much to give should be substantial enough to have them pay especially close attention to the behavior of relevant others (i.e., other's donations). Therefore, the study varied the size of (fictitious) donations of others (low/high). Based on the above, the informational social influence hypothesis would predict that subjects exposed to high donations of others would give a higher amount on the average than subjects informed of others' low donations.

Aside from its conceptual aspect, another purpose of the size-of-donation manipulation was to examine the possibility that the list technique could also be employed to increase the amounts of compliance compared with control outcomes. The key emphasis in previous compliance research centered on the proportion of subjects complying with a request. Yet, it is clear that in many everyday compliance settings, such as charity drives, a second variable of substantial practical interest is the amount of compliance. Thus, based on previous findings (Reingen in press; Reingen 1978), the high donations of other com-pliers were constructed so that they were greater than what would be normally expected for control subjects.

The study also assessed the effect the sheer number of other compliers has on a target's yielding to a request. Generally speaking, as more people agree on a given behavior, the greater the probability of informational social influence (King 1975). The informational social influence hypothesis would, therefore, suggest that subjects exposed to a long list of donors (12 donors) would be significantly more likely to contribute than subjects exposed to a short list (4 donors), and would be more likely to contribute than control subjects. Although small group research would argue that a target's likelihood of compliance increases with the number of sources of information up to the point of only four sources (Asch 1951), that research is of little relevance here because of its face-to-face interaction basis. [Evidence provided by Gerard, Wilhelmy, and Connolly (1968) suggests that when interaction is less direct, conformity tends to be enhanced as the size of the majority increases from two to seven. Since the present study involved indirect observation contexts, a number of com-pliers larger than seven was judged to be sufficient for the effect to materialize.]

When the variables size-of-donations and number-of-donors are factorially combined, the above suggests that the total compliance effects (i.e., total donations) would be strongest in the high donations/long list cell and weakest in the low donations/short list cell.


A further major purpose of the field experiment was to incorporate into the experimental design a procedure based on normative social influence so that the differential effects of both types of social influence (informational and normative) could be assessed as compliance induction techniques. [The following discussion is largely based on King (1975).] Informational social influence is largely self-inflicted and occurs when another's behavior is attended to and has impact as the result of the receiver's attempting to be "correct" or "appropriate,'' e.g., when a person's giving to charity is primarily a function of his perception of the giving behavior of others. Generally, informational social influence is unintended by the source, as is presumably the case in the present study where the fictitious donors are hypothesized to be the primary source of influence.

In contrast, normative social influence is largely instrumental and results when a receiver accepts influence in order to gain some reward beyond merely being appropriate. Normative social influence is generally intended by the source who is perceived as a mediator of rewards. Frequently, however, people can influence others to do things in normative social influence situations without mediating specific rewards (or punishments). That is, people often comply with a request because they find it intrinsically self-rewarding. To achieve this form of normative social influence, a requester provides cues, or stimuli, that are designed to make salient certain pre-established dispositions acquired during socialization, such as the internalized norm of social responsibility, which direct behaviors the target finds self-rewarding. In the context of the present study, this could mean that common additions of such cues as "We need your help" or "You should give to a good cause" may be sufficient enough to elicit greater compliance. However, such simple moral exhortations have not been found to be effective (Bryan and London 1970; Darley and Betson 1973). Cues are apparently needed that lead a target into believing that his own behavior will have an important bearing on another's welfare; i.e., cues that activate the norm of personal responsibility (e.g., Schwartz 1970). This was attempted in the present study by incorporating to the experimental script a sentence that in effect told targets that their help was needed to prevent heart attacks for people they might even know. It was predicted that subjects exposed to this cue would be more likely to contribute than control subjects.

In the final four conditions, the study assessed the compliance effects of four hybrid strategies. Subjects were first subjected to the normative social influence attempt, then were shown a list of fictitious donors, with the number of donors and size of donations varied as previously described, and finally were asked for a donation. Since the hybrid strategies would appear to reap the benefits of both normative and informational social influence attempts, they were expected to produce greater compliance with the donation request than their component compliance techniques.



Subjects were 300 male students at the University of South Carolina. Only those students qualified who were walking or sitting alone along university walkways during the hours of 10:00 a.m. to 5:00 p.m., and no subjects known to an experimenter were selected.


The experimenters, three male college students, had been thoroughly instructed and they were equipped with the identification badges, information brochures and donation envelopes commonly employed in fund-raising efforts by the Heart Association. An interaction was initiated by an experimenter's introduction of himself as representing the Heart Association. After the common introductory remarks, subjects were randomly assigned to one of ten conditions, 30 subjects each, according to a pre-specified treatment schedule that varied across experimenters. The schedule was constructed so that an experimenter completed exactly 10 replications per condition. After an interaction had been completed, an experimenter contacted the next subject that qualified.

In the first condition, the request-only control, the experimenter stated, "As part of our annual campus fund-raising drive, I'm collecting money for the Heart Association. Would you be willing to help by giving a single donation?" In the next four conditions, the list-then-request conditions, a subject was first shown a list of fictitious donors with their "donations" and then was asked for a donation. Specifically, the experimenter said, "As part of our annual campus fund-raising drive, I'm collecting money for the Heart Association. (While the experimenter stated the following, he showed the subject a list of donors of both sexes with their donations.) As you can see, other students have given a donation already. (The experimenter then counted to two and continued.) Would you be willing to help also by giving a single donation?" The length of the list (short--four donors/long--twelve donors) and the size of donations (low/high) were varied. The low donations had a mean of $.25 (range $.15 to $.35) for both short and long lists, whereas the high donations averaged to $.85 (range $.75 to $.95). The names of donors were, of course, the same across the donation size factor. Thus, three identical sets, one for each experimenter, with four lists each were utilized. [The lists were constructed by asking acquaintances of the author to write down their name and a predetermined donation amount, without collecting any funds. It is extremely improbable that subjects knew any of the "donors." Subjects were of course not permitted to add their name to the list. A blank page was attached to the list for subjects that wanted their name and donation listed. Very few of the subjects expressed that desire throughout the study.]

In the sixth condition, the normative influence-then-request condition, the experimenter stated, "As part of our annual campus fund-raising drive, I'm collecting money for the Heart Association: We need your financial support so that you can help us in preventing heart attacks for people you might even know. Would you be willing to help by giving a single donation?"

The final four conditions, the hybrid strategy conditions, were the same as conditions two through five, except that subjects were shown the lists of donors after they had been told that their financial support was needed.




Table 1 presents the results for frequency of donation, average amount donated, and total amount donated.

Compliance Frequencies

An initial Chi Square analysis on frequency of donation within each condition showed no significant differences between experimenters, with levels of significance ranging from .38 to .87. Hence, the compliance data were quite free of experimenter effects, and the subsequent analyses could therefore be performed on collapsed data.

The hypotheses with regard to compliance frequencies seemed most appropriately tested with a series of four planned orthogonal contrasts.

The first compared the combination of the long list conditions (conditions 3,5,8, and 10) with the combination of the short list conditions (conditions 2,4,7, and 9) and the control (condition 1). The informational social influence prediction that subjects exposed to a long list of compilers would be significantly more likely to contribute than subjects exposed to a short list or no list was clearly confirmed (X2 = 5.02, d.f. = l, ~ <.02). [All 2 X 2 contingency analyses were corrected for continuity and all directional tests were one-tailed (Siegel 1956). Since the comparisons were orthogonal, the error rates were not inflated.]

The second orthogonal comparison between the control and the combination of the four short list conditions was insignificant (X2 = 1.22, d.f. = 1, p > .10, two-tailed), suggesting that the compliance effects are obtained only if the number of other compliers is sufficiently large. Thus, when coupled with the results of the first comparison, the findings suggest a length-of-list effect.

In contrast, no size-of-donations effect on compliance rates was observed, as is attested to by the third orthogonal contrast between the combination of the low donation conditions (conditions 2,3,7 and 8) and the combination of the high donation conditions (conditions 4,5,9, and 10) which was insignificant (X2 = .02, d.f. = 1, p > .10, two-tailed).

It was expected that the hybrid strategies (conditions 7 through 10) would induce greater compliance than the list-only conditions (conditions 2 through 5). The fourth orthogonal comparison involving these conditions provided only marginal support for the prediction (X2 = 2.09, d.f. = 1, p < .08).

The hypothesis that the normative influence condition (condition 5) would produce greater compliance than the control (condition 1) could not be directly assessed by the series of orthogonal contrasts employed. Thus, this comparison was performed separately, with the unexpected result that no significant difference was found (X2 = .60, d.f. = 1, p > .10). Conditions 2 through 5 and conditions 7 through 10 were also analyzed separately in an attempt to determine interaction effects between length-of-list and size-of-donations on compliance. The resulting Chi Squares were insignificant (p's > .10), thus obliterating the need for a partitioning of variance via an arc sin transformation of the proportions (Rao 1952).

Compliance Amounts

Due to the presence of unequal cell frequencies of compliers (i.e., nonorthogonality) and two no-list conditions, the hypotheses with regard to donation amounts could not be tested by standard approaches for the factorial analysis of variance. Instead, a linear model appropriate for a 2 (donations: low/high) X 2 (length: short/long) X 2 (normative influence: absent/present) design that included a variable for the two no-list conditions (coded 1 if the response came from the control and -1 otherwise) was constructed (Mendenhall 1968). The importance of each effect was then assessed via general linear hypothesis procedures (e.g., Perreault and Darden 1975). Using regression analysis, simple models not including a particular effect were compared with complete models which included the effect. The results showed that only the size-of-donations factor had a significant effect on the data. Thus, the other factors could be ignored, and the analysis of the remaining one-factor model suggested a highly significant effect of size-of-donation of other compliers on the dependent measure (F = 17.61, d.f. = 1/168, p < .01). The overall means (Xlow = .37 versus Xhigh = .57) indicate that the results were in the predicted direction. [When an analysis of variance for a completely randomized design was performed on the data, the same overall conclusion emerged. Multiple orthogonal comparisons on the data showed that the high contribution conditions produced greater donations than the combination of the low donation conditions and the control; and that no significant difference existed between (a) the control and the low donation conditions, (b) the short list and long list conditions, and (c) the list-only conditions and the hybrid strategy conditions.]

Compliance Totals

Given this pattern of findings, it is not surprising that along the practical dimension of total funds obtained, the results are consistent with the informational social influence hypothesis. The conditions where long/high lists were involved (conditions 4 and 10) produced greater donation totals ($10.04 and $14.22, respectively) than their short/low list counterparts ($5.75 for condition 2 and $6.07 for condition 7). When compared with the control outcome, from a practical viewpoint the results are even more impressive. The long/high list-then-request condition produced a total three times as high as that of the control, and the normative influence, long/high list-then-request condition produced 4.2 times that of the control.


These findings are consistent with the informational social influence hypothesis, but they do not ultimately confirm it. Overall, the subjects were consistently more likely to comply with the donation request the greater the number of other compilers, and the subjects offered a consistently higher amount of compliance on the average the higher its level for other compliers. Nevertheless, alternative explanations may exist, especially in view of the fact that manipulation checks were not judged to be feasible in this field experiment.

The data also suggest that to achieve greater compliance a minimum number of other compliers is necessary if informational social influence is to be activated in indirect observation contexts. When the number of other compliers was small (four), the compliance increases tended to be insignificant. The suggestion that in indirect as opposed to direct observation settings more models are needed to pose a sufficient enough challenge to an observer's typical self-image of benevolence and charity appears plausible. The more direct a contact, the greater tends to be the influence potential (King 1975).

No support was found for a compliance technique based on normative social influence. Of course, only one of many possible operationalizations of normative influence was utilized, and its cues may simply have been too weak to produce the desired effects. It is entirely possible that with a different operationalization, one which would have made the norm of personal responsibility more salient, the results would have been more impressive. When used in conjunction with informational social influence, however, the overall compliance proportions were marginally more favorable, suggesting an augmented effect.

There is, of course, a practical utility to the findings as well. The findings suggest an effective approach to increasing the frequency and amount of compliance. Thus, the donation totals for the long/high list conditions were much greater compared with the control outcome, a result that fund-raisers in particular should consider of value. The technique can also be easily implemented, and it can be employed in an ethically responsible fashion with the use of real donors and donations. However, the effects have only been demonstrated in the context of a prosocial request. Although preliminary evidence (Reingen 1978) suggests that the technique may possess cross-situational reliability provided that the norms that govern compliance are not too strongly violated, further research is needed to determine the generality of the effect.


S. E. Asch, (1951), "Effects of Group Pressure on the Modification and Distortion of Judgements," in Groups, Leadership, and Men, ed. H. Guetzkow, Pittsburgh, Pa.: Carnegie Press, 177-90.

J. H. Bryan and M. A. Test, (1967), "Models and Helping: Naturalistic Studies in Aiding Behavior," Journal of Personality and Social Psychology, 6, 400-7.

J. H. Bryan and P. London, (1970), "Altruistic Behavior by Children," Psychological Bulletin, 73, 200-11.

A. Cann, S. J. Sherman, and R. Elkes, (1975), "Effects of Initial Request Size and Timing of a Second Request on Compliance: The Foot-in-the-Door and the Door-in-the-Face," Journal of Personality and Social Psychology, 32, 774-82.

R. B. Cialdini and K. Ascani, (1976), "Test of a Concession Procedure for Inducing Verbal, Behavioral, and further Compliance with a Request to Donate Blood," Journal of Applied Psychology, 61, 295-300.

R. B. Cialdini, J. T. Cacioppo, R. Basset, and J. A. Miller, (in press) "The Low-Ball Procedure for Producing Compliance: Commitment then Cost," Journal of Personality and Social Psychology.

R. B. Cialdini and D. A. Schroeder, (1976), "Increasing Compliance by Legitimizing Paltry Contributions: When even a Penny Helps," Journal of Personality and Social Psychology, 34, 599-604.

R. B. Cialdini, J. E. Vincent, S. K. Lewis, J. Catalan, D. Wheeler, and B. L. Darby, (1975), ,'Reciprocal Concessions Procedure for Inducing Compliance: The Door-in-the-Face Technique," Journal of Personality and Social Psychology, 31, 206-15.

J. M. Darley and D.C. Batson, (1973), "From Jerusalem to Jericho: A Study of Situational and Dispositional Variables in Helping Behavior," Journal of Personality and Social Psychology, 27, 100-8.

M. Deutsch and H. Gerard, (1955), "A Study of Normative and Informational Social Influences upon Individual Judgement," Journal of Abnormal Social Psychology, 51, 629-36.

J. L. Freedman and S. Fraser, (1966), "Compliance without Pressure: The Foot-in-the-Door Technique," Journal of Personality and Social Psychology, 4, 195-202.

H. B. Gerard, R. A. Wilhelmy, and E. S. Connolley, (1968), "Conformity and Group Size," Journal of Personality and Social Psychology, 8, 79-82.

S. W. King, (1975), Communication and Social Influence, Reading, Mass.: Addison-Wesley Publishing Company.

W. Mendenhall, (1968), The Design and Analysis of Experiments, Belmont, Ca.: Wadsworth Publishing Company.

C. W. Park and V. P. Lesig, (1977), "Students and Housewives: Differences in Susceptibility to Reference Group Influence," Journal of Consumer Research, 4, 102-10.

W. D. Perreault and W.R. Darden, (1975), "Unequal Cell Sizes in Marketing Experiments: Use of General Linear Hypothesis," Journal of Marketing Research, 12, 333-43.

C. R. Rao, (1952), Advanced Statistical Methods in Biometric Research, New York, N.Y.: Wiley.

P. H. Reingen, (1977), "On the Marketing of Fund-Raising Campaigns: Foot-in-the-Door and Legitimization of Paltry Contributions," Working Paper, Center for Marketing Studies, University of South Carolina.

P. H. Reingen, (1978), "Social Influence as a Compliance Induction Technique," Working Paper, Center for Marketing Studies, University of South Carolina.

P. H. Reingen, (in press), "On Inducing Compliance with Requests," Journal of Consumer Research.

P. H. Reingen and J. B. Kernan, (1977), "Compliance with an Interview Request: A Foot-in-the-Door, Self-Perception Interpretation," Journal of Marketing Research, 14, 365-69.

S. H. Schwartz, (1970), "Moral Decision Making and Behavior,'' in Altruism and Helping Behavior, eds. J. Macaulay and L. Berkowitz, New York: Academic Press, 127-41.

C. A. Scott, (1976), "The Effects of Trial and Incentives on Repeat Purchase Behavior," Journal of Marketing Research, 13, 263-69.

C. A. Scott, (1977), "A Self-Perception Approach to Modifying Socially-Conscious Behavior: The Foot-in-the-Door Technique," Journal of Consumer Research, 4.

S. Siegel, (1956), Non-Parametric Statistics for the Behavioral Sciences, New York, N.Y.: McGraw-Hill.

C. Wagner and L. Wheeler, (1969), "Model, Need, and Cost Effects in Helping Behavior," Journal of Personality and Social Psychology, 12, 111-16.