Cann, Sherman, & Elkes, 1975

Study 1

Study 2

95% of subjects agreed to the initial request, regardless of its size, and responses did not vary with sex of requester. In the cases of the second, larger request, there was no significant difference between the 3 groups when visited by a female requester. Subjects visited by a male requester had a higher percentage of compliance when asked the initial small request and then the later, larger request. So, it appears that the foot-in-the-door effect occurred among subjects in those groups with male requesters/initial small request. Initial request was only succesful in getting compliance to the second, larger request if it was of small magnitude.

Results showed that the subject perception that others had complied before(reactance) had little effect on whether or not a agreed to a request. Over all conditions, 27% of the subjects who complied with the first request also complied with the second larger request, offering support for the foot-in-the-door effect. Only 17% of the control condition complied with their single, large request.

Contrary to the hypothesis, as perceived prior compliance increased, compliance to the second request decreased. Also, as reactance level increased, subsequent compliance to the second request decreased.

The prediction that the extreme-then-critical condition would cause more behavioral compliance than the minimal-then-critical conditionwas support ed. Also the critical request only control condition showed greater behavioral compliance- gaining than the minimal-then-critical condition at a significant level.

The prediction that extreme-then-critical condition would show more subsequent help compliance than the other two groups was strongly supported. Extreme-then-critical= 6/7; minimal-then-critical= 0/2; and critical request only group= 3/7.

Dejong, 1981

Dejong & Funder, 1977

There were no sex differences for verbal compliance. Fifty-five percent of males and 53 percent of females verbally agreed to donate blood. There was no difference for behavioral compliance, with 27 percent of males and 25 percent of females actually appearing at the donation center. There was no variation in compliance across conditions and the differences were not significant. There was practically no variation in verbal compliance a cross conditions. Behavioral compliance was substantially less than verbal compliance with only about half of those who verbally complied actually showing up to donate blood. Information about previous donation history was obtained in order to explain differences in behavioral compliance of those who verbally complied. Eighty-two percent of those people who had donated before carried through their verbal commitment, whereas only 21 percent of those who had never donated did so. Thirty-seven percent of the variation in the behavioral compliance in this group was accounted for by donation history. It is evident that the relationship between donation history and behavioral compliance holds up within both the experimental condition and the control condition. This suggests thst the experimental treatment had no effect on the relationship between the variables.

Foss & Dempsey: Study 2, 1979

In this study, 25 persons could not be recontacted for the critical request. Also, persons who said they might donate were classified as refusals. People who were previous donors were removed from the analysis. The greatest verbal compliance was obtained in the large initial request condition, but in neither case was this a significantly different from the control condtion. By combining the experimental groups, the overall compliance rate was 26% which does not differ very much from the 20% compliance obtained in the controlled condition. There was no behavioral compliance in this study. Only four donated blood and one of those four had not been successfully contacted for the crirical request.

Foss & Dempsey: Study 3, 1979

One apparent foot-in-the-door effect emerged in this study, but it was more likely do to some other factor than the experimental treatment. There were some moderate variations in verbal compliance, but these did not approach statistical significance. Compliance was greatest in the poster conditions, but did not differ significantly from the control condition. The 40% compliance rate among combined experimental conditions did not differ significantly from the control group. For the entire sample, behavioral compliance in the experimental condition was somewhat higher than in the control condition and the overall effect approached conventional levels of statistical significance. The poster condition accounted for this effect was significantly different from the control condition whereas the other two experimental condition did not differ significantly from the control condition. Donation history was obtained from the participants in order to see if it would play a strong role in determining which people of those who agreed to donate blood would actually do it. Forty-eight percent of those who agreed to donate were previous donors, and of those 39% actually donated. Twenty eight percent of those who agreed to donate, but had not donated previously, actually did so. Donation history is still important determinant of behavioral compliance.

Freedman & Fraser Study 1, 1966

Apparantly even the small request was not considered trivial by some subjects. Only about two thirds of the subjects in the Performance conditions and Agree-Only conditions agreed to answer the questiones about the household soap. However, some of the subjects who refused the first request later agreed to the large request. Over 50% of the subjects in the Performance conditions agreed to the larger request, while less than 25% of the One-Contact condition agreed to it. In this study, it appears that compliance with the small request does increase compliance with the large request.

Freedman & Fraser Study 2, 1966

There was no large difference among the experimental conditions as far as agreeing to the first request. The first request tended to increase the degree of compliance with the second request. Fewer than 20% of the controls agreed to put a large sign on their lawn. Over 55% of the experimental subjects agreed with over 45% being the lowest degree of compliance for any experimental condition. As expected, those conditions in which the two requests were similar in terms of either issue or task produce significantly more compliance than the control conditions.

Fish & Kaplan, 1974

Furse, Stewart, & Rados, 1981

Hansen & Robinson, 1980

Rosponse rates were greater for the four foot treatments groups than for the corresponding control groups. The control groups were significantly slower in responding to the mail survey than the four telephone contact groups. The results indicate no relationship between foot manipulations and response completeness.

Harris, 1972

All subjects who were asked for directions gave them. All subjects wearing watches gave the time exept one. Only 11.11% of the subjects in the Dime Only condition gave the experimenter a dime, but 44.4% of the subjects in the Time condition and 38.9% of the subjects in the Directions condition gave the experimenter a dime. There was no significant difference in the number of dimes given by women and men.

Harris, Liguori & Stack, 1973

The precentages of the subjects offering to contribute in the Favor, Bribe, Control 1, and Control 2 conditions were 30%, 26%, 25%, and 25%. There was no significant differences among the four experimental conditions or between the two control conditions. There were no significant experiment er, sex, or ethnic effects. Fifty-eight of those who participated in the experiment refused to grant the small favor.

Chi square analyses show that there was a significant difference in percentages of people giving money in the five theft conditions after the first and second requests, Chi square (4) = 14.98, p < .01and Chi square (4) = 16.30, p < .01 respectively.

The percentages of subjects donating in the Stop Thief 15%, Thief 52.5%, and Control 25% conditions are significantly different after the first request, Chi square (2) = 13.49, p < .01, similarly the corresponding percentages of 30%, 70%, and 35% after the second request are also significantly different, Chi square (2) = 14.29, p < .01.

The analysis of variance on the amount give n after the first request revealed that subjects gave more money in the High Need than in the Low Need conditions. There was a significant theft effect , but no significant interaction. Those subjects in the two Thief condition s donated more than those in the Control and the Stop Thief conditions.

The analysis of variance showed similar findings when the amount of money given after the second request was analyzed. Subjects gave more money to a High Need than to a Low Need request. There was a significant theft effect, but no significant interaction. Subjects in the Thief conditions gave more than those in the Control and the Stop Thief conditions. However, those in the two Thief conditions and those in the two Stop thief conditions did not differ significantly.

Experiment 2

The percentages of subject s giving money after the initial request did not differ significantly. Chi square (4) with Yate's correction = 1.35 and 3.06, respectively, p > .05. The analysis of variance of amounts of money given show a significant treatment effect after the second request, F = (4,65) = 2.54, p < .05, but not after the first request, F = (4,65) = 1.34, p > .05. Subjects in the N o Thief Same condition were significantly more generous than those subjects in the other four groups combined, F = (1,65) = 4.84,p < .05. Those subjects in the Stop Thief Same condition followed suit as having the next highest generosity being not significantly less altruistic than the combination of the other four groups, F = (1,65) =2.27, p > .05.

Looking at the data on amount of money donated the results are similar. Subjects in the no prior condition donated on the average $.58 , while those in small and moderate conditions donated $.98 and $.87, respectively. Prior requests made a significant difference in amount of money donated. No prior request condition ( F = (1,85) = 5.15, p < .05) the small and moderate conditions did not differ.

Due to the greater number of subjects in the small and moderate prior request conditions that donated more and on average those conditions donated more money. In order to account for the various conditions that could make up these findings, nondonors were eliminated and the contributions of only donors were compared . After analyzing this data the three groups did not differ at all. This tells us that prior request seemed to have an all-or-none effect on the probability of a donation rather than a moderating effect on the size of a donation once a subject decided to contribute.

The preliminary chi square test on frequency of donation within each condition yielded no significant differences due to sex of subject with levels of significance ranging from .28 to 1.00. Levels between .36 and .89 showed no significant differences between experimenters.

It was predicted that experimental conditions would produce greater compliance gains than the direct request control conditions. This was confirmed, chi square = 5.115, df =1, p = .024.

The next prediction was that adding the even a penny will help catch phrase would produce greater compliance than just the small-then-donation request and extreme-then-donation request conditions. This received only directional support, chi square = 2.00, df = 1, p = .16.

Results indicate that using just the direct donation request and the penny catch phrase made it very hard for subjects to refuse when comparing this condition to the other four experimental group by orthogonal comparison. Chi square < 1, df = 1, ns.

In comparing the small-then-request and the extreme-then- request conditions with the small-then-request, even-a-penny and extreme- then-request conditions they too were found to be insignificant. Chi square < 1. df = 1, ns.

Analyzing the mean donation for complying subjects there was no significant difference predicted. Utilizing the 3X2 ANOVA the results supported the null prediction. Neither main effects F =1.27, df = 1/67, p = .26 nor interaction effects for the even-a-penny catch were significant.

The even-a-penny condition produced 1.5 more donation than the direct request.

The extreme-then-request, even-a-penny produced 1.8 times more than the direct request.

The small-then-request produced 2.1 times more than the direct request.

The extreme-then-request produced 2.2 times more than the direct request.

The small-then-request, even-a-penny produced 2.3 times more than the direct request.

The following are the results of the Volunteer request condition
: The extreme-then-request yielded zero compliance with being a future helper.
The volunteer request only control condition yielded two subjects.
The small-then-request yielded two subjects.
The small-then-request, even-a-penny yielded three subjects.
The extreme -then-request, even-a-penny yielded three subjects.
These results indicate that the orthogonal comparison of volunteer-request only control condition versus the other four experimental conditions was insignificant and was contrary to the predictions.

Replication

Orthogonal comparison of experimental groups to control groups produced the following results: Chi square = 4.67, df = 1, p = .03. This tells us that the experimental groups were greater in gaining compliance than the control groups.

Orthogonal comparison of small-then-request, even-a-penny and the even-a-penny group produced the following results: Chi square < 1, df = 1, ns. This tells us that there was no significant difference between the se groups.

The ANOVA on mean donation of complying subjects showed no significant difference. F < 1. The even-a-penny condition produced three times that of the control, while the small-then-request, even-a-penny produced four times that of control group donations.

The results of the initial analysis examining the effect of the manipulation on the dependent variables was a linear function discriminating significantly between the control and experimental conditions. The linear function weight was .67 for the behavioral measure of helpfulness, .46 for self-perception of helpfulness, and, -.61 for potential unpleasantness in helping situations.

The significant results of the discriminant function analysis were follow ed by one-tailed t tests on the individual dependent variables. The results were in the predicted direction for each of these variables. The behavioral measure increased significantly from .50hr to .93hr, for the control versus experimental groups.

The third dependent variable on external attributions concerning the help ing situation showed a significant decrease from 5.9 to 4.0.

For the 29 subjects, self-perception of helpfulness had a correlation of r=.18, with the amount of time volunteered. The perceived threat on helping situations, had a correlation of r=.01 with the amount of time volunteered. The absolute level of the perceptual measures did not have a strong correlation with the amount of helping behavior.

The number of subjects who volunteered a large amount of time for the second request was 3 of 15 (20%) for the control condition, and, 8 of 14 (57%) for the experimental condition.

The results sugest that success of the foot-in-the-door technique depends to some extent on the size of the first request. The two smallest sized first requests were ineffective in producing future compliance, where as, the two largest sized first requests were significantly effective.

The results of this research experiment were that no differences were found between the five experiments in obtaining compliance with the first request, nor were any significant differences observed between those who received the attribution measure and others for either the second request behavioral intentions, or behavior. Thus, respondents were pooled within experimental treatments for subsequent analysis.

Positive behavioral intention to comply with the initial request, i.e., verbally agreeing to, and actually placing the sign in the window, was 76.7%, with no incentive, 65.6% with a $1 incentive, 72.9% with a $3 incentive, 74.1% with a double incentive. Resulting in an overall intention toward compliance with the initial request of 72%,(207 respondents).

Compliance behavior, with the first request was, 47.8% with no incentive, 55% with a $1 incentive, 65.1% with a $3 incentive, 55% with a double incentive, and 55.7%(149) overall behavioral compliance. Positive behavioral intention to comply with the second moderate request with no incentive was 45.2%, with a $1 incentive 46.7%, with a $3 incentive 37%, with a double in centive 60%.

Positive behavioral intention to comply with the second larger request, with no incentive was 37.9%, with a $1 incentive was 16.1%, with a $3 incentive was 37.5%, with a double incentive was 33.3%. The actual behavior compliance was 31% with no incentive, 12.9% with a $1 incentive, 28.1% with a $3 incentive, 25% with a double incentive.

There was no difference between the effectiveness of foot-in-the-door and persuasion in both experiment 1 and experiment 2.

A much larger proportion of normal weight subjects volunteered when the questionnaire was distributed prior to the solicitation (68% to 31%). However, there was little difference between the compliance of the overweight participants (61% of the questionnaire first compared to 54% of the solicitation first).

Subjects in the no-pay condition were more likely to comply with the second request than were subjects int he control group. Subjects in the pay condition were less likely to comply with the second request than were subjects in the control condition but the difference was not significant. Rate of compliance in the no-pay condition was significantly higher than rate of compliance in the pay condition.