Navigation – Plan du site

Integrating Exchange and Heuristic Theories of Survey Nonresponse*

John Goyder, Luc Boyer et Guil Martinelli
p. 28-44

Résumés

L’intégration des théories d’échanges et des théories heuristiques de la non-réponse dans les enquêtes : Depuis dix ans, on constate un intérêt croissant dans la théorisation de la non-réponse dans les enquêtes: Des chercheurs tels que Dillman, Groves et Couper ont construit pour tous les domaines des sciences sociales des schémas conceptuels pour expliquer pourquoi des gens acceptent ou refusent de participer à des enquêtes. Mais l’approfondissement conceptuel est source de complexification et de confusion. En débutant avec l’analyse de l’usage actuel, dans des articles sur la non-réponse dans plusieurs revues reconnues, des relations entre la théorie des échanges et la théorie des heuristiques psychologiques en faveur de la participation, nous construisons un schéma conceptuel à deux dimensions. Basé sur des travaux de Groves et Couper, une dimension est définie comme « l'importance de la prise de décision ». Ceci aide à situer l’interprétation heuristique des décisions de participer. La seconde dimension, appelée « force des facteurs culturels », clarifie les variétés différentes de la théorie des échanges qui examinent la décision de participer. De ce point de vue, il devient clair qu’une source de confusion dans la théorisation de la non-réponse est l'hétérogénéité des approches rassemblées sous le nom « d'échanges sociaux ». Au niveau le plus général, les échanges sociaux forment l’arrière plan de la plupart des théorisations de la non-réponse, mais une fois définis plus spécifiquement, il est possible de les emboîter dans l’une ou l’autre des cases d'un tableau à deux dimensions.

Haut de page

Notes de l’auteur

* This is an evolution from a paper presented by John Goyder, Keith Warriner, Susan Coutts (now Miller), and Maryanna Basic, "Socio-Economic Bias in Surveys: Some Attempts at Interpretation," at the International Conference on Survey Nonresponse, Portland, Oregon, 1999. It is a further revision of a paper presented at the 60th annual AAPOR Conference, Miami, Florida, May 12-15, 2005, session on "Theories and Models of Nonresponse."

Texte intégral

Introduction

  • 1  Throughout this discussion we are setting aside the issue of achieving contact with sample listing (...)

1A generation ago, it was reasonable to comment on the atheoretical nature of research on why people do or do not respond to requests for surveys.1 Hawkins (1975:461), Linsky (1975:101) and Cox (1976:104) were among those in the 1970s noting bias toward technique over theory. The theory-vacuum began to fill with Dillman's (1978) widely noted Mail and Telephone Surveys. Survey behaviour was presented as an instance of social exchange (see also the Dillman-supervised Master's thesis by Gallegos, 1974). Dillman argued that diffuse, social rather than purely economic, calculations of costs and benefits take place when people decide about participating in a survey. A decade later, Goyder (1987) suggested that seemingly atheoretical research reports carry embedded assumptions which are the basis for latent theories of nonresponse. The Silent Minority thus argued that methodologists who disavowed any theoretical agenda were often implicit behaviourists who reflected a Skinnerian view of human behaviour, still influential when much of the first generation of research on survey nonresponse was being conducted.

2Important new theoretical contributions began with Groves (1989), followed by his widely-cited Public Opinion Quarterly piece (Groves et al., 1992). This drew attention to psychological knowledge about compliance-eliciting as relevant to response decisions. From this perspective, behaviours in reaction to requests for surveys are not necessarily determined by a considered process of social exchange decision-making around costs and benefits. In the case of telephone calls made "cold," without prior notification, participation or non-participation may be resolved within a few seconds, a point Dillman (1978:17) had noted also. People may fall back on short-cuts to decision making; that is, cognitive heuristics. The mind classifies the request as belonging to a category of events, and, from this act of classification, the decision about participation follows. Factors such as immediately liking the fieldworker or sensing that the interview is a scarce opportunity may aid the decision. The heuristic approach shaped the conceptual orientation to Groves and Couper's (1998) study of nonresponse on government-sponsored personal interview surveys.

3Early in the new century, the relationship between exchange and heuristic theories is uncertain. Dillman (2000:15-21), in his extensive revision of the 1978 classic, treated principles of psychological persuasion as a sub-theme within social exchange theory. Social proof or validation -- for instance, the notion from Cialdini (1993) that knowing other people have taken an action makes oneself more likely to do the same -- was fitted into the social exchange framework as a form of reward for respondents. Groves and Couper (1998:32-33), in contrast, had bundled social exchange, at least in part as distilled down to the norm of reciprocity, as a sub-theme within the heuristics articulated by the psychology of compliance. In a jointly edited collection from 2002, Groves, Dillman et al. presented opportunity cost, topic salience, and social exchange as separate elements under the heading "influences on survey participation," and yet a wider interpretation of social exchange (e.g., Dillman, 1978) would see opportunity cost and salience as part of the social exchange package. Stoop (2005:101) notices the lack of theoretical integration in The Hunt for the Last Respondent, referring to "different nonresponse school(s)." There is, then, some conceptual inconsistency. Should exchange theory and heuristics theory be regarded as competitors or as complements? It is the kind of figure-ground dilemma noted by Marshall McLuhan in a communications theory context (Gordon, 1997:149). Between social exchange and psychological heuristics, which is the background and which the figure?

4In the pages below, we shall attempt to resolve this issue, and in so doing propose resolution to two controversies within survey methodology.

Theories of Survey Nonresponse in Use

5Menzies (1982) discussed "sociological theory in use," examining the "research theory" which people conducting empirical studies use. We borrowed this idea and began by looking at of theories of survey nonresponse in use, for the purpose of establishing which approaches are dominant and how the fit between theories is presented. Menzies's (1982) notion about "in use" was that theory as deployed by empirical researchers is not necessarily the same as theory from textbooks. Theoretical concepts can drift and be adapted, often without explicit re-formulation. The theory-in-use approach thus provides a reality test of our depiction, above, about current theorizing around survey nonresponse. Is it correct to say that Dillman and Groves are the two dominant theoreticians here, and that their approaches talk past each other? The analysis next described addresses that. Articles in seven journals which frequently publish in the survey methodology area were examined, over the period 2002-2005. These were Public Opinion Quarterly, Bulletin de Méthodologie Sociologique, Journal of Official Statistics, Quality and Quantity, Field Methods, the International Journal of Public Opinion Research, and Sociological Methods and Research. Restricting the analysis to 2002 and later gave the cognitive heuristic theory time to settle into the literature following its wide exposure resulting from the exposition in Groves and Couper (1998).

6Of the scores of articles examined, 47 were deemed relevant to an analysis of theory-in-use around survey nonresponse. Only research reports, not commentaries or expositions, were included. Studies just on item nonresponse were excluded, as was work examining the frequency distributions on questions from different kinds of experimental condition (e.g., question layout, formatting, and fonts).

  • 2  It cannot be claimed that this coding is an entirely scientific, objective exercise that another o (...)

7The theoretical approach toward survey nonresponse used by the author(s) was coded, concentrating on the initial perspective, not the amended conceptual conclusions which might be reached after a presentation of empirical data.2 For example, Teitler et al. (2003) refer to recent studies finding nonresponse to be ignorable. Our theory-coding here was that the implicit conceptual scheme was nonresponse as random event even though in their concluding discussion the authors began to depart from a random nonresponse interpretation. There were 14 instances of unstated yet implied theory, and 63 more explicit references to theories about nonresponse. In counting totals, the distribution for which is shown in Figure 1, the implicit theories are weighted at 0.5 instead of 1.0.

Figure 1:

8Although social exchange is the leading perspective in the journals examined, another clear impression is the wide variety of ideas guiding empirical studies in nonresponse. The cognitive heuristic approach is the fourth most commonly used, behind "social context/social disorganization" and then socio-demographics. Context includes determinants such as time of year, kind of neighbourhood being surveyed etc. Socio-demographics affect the "householder's psychological dispositions" (Groves and Couper, 1998:32), indicate likely salience, and hold yet other meanings. Here and elsewhere, questions about the correct "nesting" thus arise. In other words, are there here a dozen or so distinct approaches, or do some fit within others. "Emotion" as a factor in response decisions (e.g., Kaplowitz and Lupi, 2004:199) in some respects nests within heuristics, and "attitude" seems to fit into a broadly-conceived social exchange framework. As treated by Stock‚ and Langfeldt (2004:7) for example, attitude is compatible with "instrumentally rational meta-decision." "Minimization of burden-intrusion" has elements of both exchange and heuristics. Consider, for instance, the research note by Rizzo et al. (2004:267) in which is discussed ". . . a new and simple method of within-household sampling that is minimally intrusive and, thus, should not have a deleterious effect on response rates." There are hints of both exchange and heuristics being on the authors' minds when they note that the Kish question must be asked "before the interviewer has had time to establish a relationship with the household respondent."

9An example of a theoretical approach well outside the dominant heuristics/exchange axis is health and mental ability of the person sampled, handled perhaps most interestingly in Hauser's (2005:27) observation that cognitive functioning has been neglected in theorizing on survey nonresponse.

  • 3  As Groves and Couper (1998: 130) observed, exchange theory is very "diffuse," but also their data (...)

10Little interlocking occurs among the theoretical approaches. Of the 47 articles, 22 (including implicit-theory scores) cited just one theory; 20 cited two; and in the remaining five articles, three theories were cited. In just five of the 47 articles (four explicit, one implicit), heuristics and social exchange were cited side by side and none of these explained how the two theories relate to one another. In no instance was an attempt made to deduce alternative predictions from heuristics and exchange and test them with data. The closest example would be from outside our data base, from Groves and Couper (1998:125-131) in a discussion about response decisions by low SES people on US government surveys.3 Our initial sense of an un-integrated theoretical landscape around survey response seems to fit.

Social Exchange and Cognitive Heuristics

11The reason for uncertainty as to how exchange and cognitive heuristics relate derives from the way that exchange is not one but several only loosely unified conceptual approaches. Social exchange theory performs a "sensitizing function" (Gergen et al., 1980:270). It is "not a theory at all. It is a frame of reference within which many theories . . . can speak to one another. . ." (Emmerson, 1976: 336). It is an "umbrella" and a "plurality" (Goyder, 1987: 165-6), with "many shapes and sizes" (Heath, 1976:170, although for attempts at synthesis see Makoba, 1993 or Zafirovski, 2003, 2005). Social exchange in its broad sense suggests a continuum, ranging from decisions based on habits and scripts right up to the deep weighing of every alternative. Social exchange and heuristics are thus not mutually exclusive. Brehm (1993:82-83), indeed, argues that heuristics are best at explaining refusals while rational calculation of costs and benefits best accounts for completions.

12When social exchange is considered at the level of sub-types, it becomes clear why the overlap between exchange and heuristics is only partial. We will identify four sub-types of social exchange, to suggest some of the different forms and features of exchange that bear on survey participation. First, social exchange as developed by the micro-theorist George Homans presents exchange as overlapping with psychological heuristics. The following passage from Homans (1958:602), for example, reproduces the "liking" heuristic from Groves and Couper (1998): "If you think that the members of a group can give you much- - in this case, liking- - you are apt to give them much- - in this case, a change to an opinion in accordance with their views." The change of opinion the survey fieldworker seeks is, of course, a reluctant respondent's decision to participate in a survey. Admittedly, a survey interview is not a "group" in Homans's sense. We are not saying that psychological heuristics fit perfectly with exchanges from the small groups research analysed by Homans, but some key assumptions are shared. Cialdini's (1993) book on persuasion identifies some of the exchange aspects of heuristics. He begins by noting the behaviour of mother turkeys, who make decisions entirely on whether a living thing sounds like a turkey chick or not- - the power of a heuristic. Repeatedly in the rest of the book, however, Cialdini (1993: 156) outlines the ways in which humans subjected to heuristic scripts differ from animals. Humans "counterattack" against heuristics by learning from experiences. They act to avoid forced reciprocation and use cultural "jujitsu" as a means of "how to say no."

13Anthropologists have inferred forms of social exchange rather different from that analyzed by Homans, much more culturally rooted, but still consistent with heuristic decision-making. In the Kula exchange studied by Malinowski, for example, Trobriand Islanders performed exchanges of necklaces and armbands in which societal as well as individual functions were at work. The exchange was symbolic and important for the fostering of trust and good relationships between tribe members. Malinowski (e.g., 1926:25) frequently refers to the "ceremonial" dimension to the exchanges, and ceremony surely is a form of heuristic (Camic, 1986: 1052). Of course it makes sense to place tribal exchange as low in decision-making only insofar as late-modern Western society is the context for comparison. It would misrepresent Malinowski to portray the islanders as unreflecting slaves to their culture (Kuper, 1983: chapter 1), but in comparative context this was a people for whom many decisions were culturally prescribed. The reciprocity heuristic, for example, was strong among these islanders (Malinowski, 1926:23-24). Homans was looking for general behaviouristic principles holding across all cultures; the Kula exchange was highly defined by the culture, yet both used a "shallow" form of decision-making. As Groves and Couper (1998:32; see also Brehm, 1993:83) have said with reference to survey response decisions:

(m)uch social and cognitive psychological research on decision making . . . has contrasted two types of processes. The first is deep, thorough consideration of the pertinent arguments and counterarguments, of the costs and benefits of options. The second is shallower, quicker, more heuristic decision making based on peripheral aspects of the options.

14While the social exchange seen by both Homans and Malinowski involves the heuristics of shallow decision-making, two other forms do not. Consider Peter Blau's analysis of a form of exchange inferred from a case study of employees in a bureaucracy. These were reflexive, decision-making actors working within a well-defined office culture. The exchanges took place within on-going relationships in which trust (see Dillman, 1978:16 and 2000: 19 for the importance of this term for survey behaviour) was high because a rule was being broken by the exchanges, leading to "emergent social structure" (Blau, 1963: 127-8; 1964:3). Survey participation does not perfectly fit Blau's model of exchange. Only in the most extensive of panel studies would the survey relationship approach the level of relationship the federal bureaucrats held with one another. Many of the social components do apply, however, such as the need to control the number of requests (Blau, 1963:130), anxiety over rejection (1963:131), and the trust factor already mentioned. Dillman's (1978) Total Design Method (TDM) and attendant theorizing about the social exchanges involved in survey participation fit here. By careful application (Dillman et al., 1984) of TDM techniques, a balance sheet in favor of participation accumulates bit by bit. Small prepaid cash incentives are usually theorized in Blauian social exchange terms, drawing on an argument about the cultural force of reciprocity (e.g., Warriner et al., 1996:546). Gouldner (1960), in a piece frequently quoted in reports on incentives experiments, saw this as a core norm of most societies. The prepaid cash incentive, most often used in the self-administered modes, in which more than one contact is typically made, and using a medium such as paper or the Internet, allows the potential respondent time to mull over his or her decision and let the reciprocity norm percolate along with other considerations such as salience of topic.

15A fourth form of social exchange, rational-choice exchange, is well expressed within survey response-maximizing technology by large payments, which are usually delivered post-completion. This is deep decision-making not necessarily very culturally prescribed. There are frequent examples in the literature of surveys of high-earning occupations such as physicians in which participation is encouraged by payments of $50 and up (e.g., Mizes et al., 1984; Berry and Kanouse, 1987; or more recently, Halpern et al., 2002). Here, the exchange has shifted away from the social-cultural realm. It is on the outskirts of social exchange as many survey methodologists would understand the term (e.g., Stoop, 2005: 90). The phrasing used in the Gunn and Rhodes (1981:112) survey of physicians paid experimental amounts of $25 or $50 in 1979 dollars sets the economic foundation of the exchange: "Dr. ____ will be paid $____ for the time he spends on the interview." As Berry and Kanouse (1987:104) argue, even here there may be a symbolic component to the payment, but still the balance has shifted away from the logic behind small prepaid incentives.

16These arguments about heuristics and types of exchange are summarized in Figure 2, using as horizontal axis, "amount of decision-making," as described above. "Deep" decision-making admittedly may be more typical of institutions than of individuals. It is a deep, or "aided-analytic" strategy, for example, when the route for a new highway is selected. All kinds of data, modelling and consultation would go into such a decision. Hanging up on a telemarketer, in contrast, based simply on the tone of voice of the caller, is the result of a shallow or "nonanalytic" strategy. Cognitive scripts and homiletic rules will dominate the decision to participate. Medium or "unaided-analytic" strategies (Beach and Mitchell, 1978) rest between heuristic assessments and deep evaluation of costs and benefits of answering a survey. Here, the potential respondent assesses response alternatives using the "satisficing" criterion or using a small subset of the information presented (Beach and Mitchell, 1978:441-443). As Groves et al. (2000) have shown with their concept of leverage-salience and thresholds around response behaviour, a salient topic might induce potential respondents to engage in "extensive cognitive processing of the message itself", whereas a non salient topic might sway them towards "peripheral cues to make a judgement" (Groves et al., 2004:4). The acknowledgment of two types of decision-making processes as bases for answering a survey thus does not bind one to an irreconcilable dichotomy between heuristics and rationalism. Social exchange means bounded rationality. It straddles a continuum, ranging from decisions based on habits and script theory right up to the fully reflective person weighing every alternative.

Figure 2:

17The vertical axis, "strength of cultural factors," draws attention to the importance of culture as an essential dimension both of social exchange and cognitive heuristics. Following Befu , the motivation for, and nature of, social exchanges acquire their meanings once situated in a cultural context: "the socio-cultural context sets the stage on which participants can act out their exchange behaviour according to 'the script', i.e. the dictates of the exchange model. Without the stage, actors cannot act; without the cultural context, exchange analysis is vacuous" (Befu, 1977:259). This is not to imply that cultural contextualism is synonymous with determinism. As Wrong (1961) pointed out, sociologists are prone to an oversocialized conception of human beings. Norms and values simply serve an orienting function for an individual socialized in a given culture.

18In the survey context, cultural explanation was articulated by Johnson et al. (2002:55) when they proposed that "(s)urvey respondents are not merely autonomous information processors, rather, they all exist within complex social matrices that influence their thoughts, feelings, and behaviours." They saw the respondent's culture having an impact on both communications styles, social participation and socio-economic opportunities (Johnson et al., 2002: 68). Cultural difference is implied within cross-national comparisons of nonresponse rates (de Leeuw and de Heer, 2002) and was important for Durand et al. (2002) with reference to the greater success of Quebec pollsters compared with those in English-speaking Canada.

19The strength of cultural factors on the decision to participate in a survey will depend upon the particular circumstances which the individual may be facing. This level of uncertainty reflects the fact, according to Befu, that much of the cultural norms of behaviour leave details governing social exchange unspecified. The norm of reciprocity, crudely operationalized as a prepaid incentive, does not specify by itself cultural expectations surrounding the amount or the manner in which such a payment should be made in return for the cost of filling out a survey. Such details are to be found, following Befu, in the rules of exchange and the strategies of exchange. The rules of exchange will specify "a range of possibilities as to what may be given and how it may be given," the strategies of exchange will determine "the exact content of the resource to be given, its quantity, and the manner of prestation"; at base, however, "these decisions are often, if not always, made on the basis of the individual's desire to maximize the opportunity for his benefit/profit" (Befu, 1977:259-260). The net result of norms, rules and strategies is a culturally contextualized but highly dynamic and localized form of exchange. As a consequence, cultural factors such as norms and values will exert a diffuse and variable amount of pressure on social exchange relations.

20To demonstrate this variability, we have labelled the vertical axis from strong to weak, contrasting explanation in terms of culture against individual-internal explanations such as personality or primal calculation of self-interest. Cultural factors will be strong when norms, rules and strategies are aligned with well-trodden social conventions such as responding to a government census or reciprocating to a small prepaid incentive. Two instances may account for weak factors: first, if norms and rules about survey response are unknown, imprecise or unavailable such as in the case of survey inexperience or novel survey practices; second, cultural factors may be subverted by principled reasons, self-interestedness, and egoistic personality traits. Having little discretionary time (Groves and Couper, 1998:122), for example, may overrule other normative guidelines such as "being helpful." Lastly, medium cultural factors suggest exchange relations governed by social psychology where personality traits and cultural factors intermix.

21Cognitive heuristics straddle the strength of cultural factors vertical dimension, as the oval in Figure 2 indicates, but are clustered at the shallow decision-making end of the horizontal axis. A decision strategy based on social validation, for example, a heuristic noted by Groves and Couper (1998:33-34) is cultural. "If householders believe that most people like themselves agree to participate in surveys, they may be more inclined to do so themselves." We say this heuristic is culturally variable with Riesman (1950) -- inner-directed vs. other-directed national character -- in mind. Thus, although the concept of heuristics originates from psychology, the content is sometimes sociological-cultural, as the tribal exchange example illustrated. For another example, the principle of reciprocity, frequently invoked both by exchange and by heuristics theoreticians as already noted, is an example of a determinant of response behaviour which can operate at the heuristic level but draws from culture. Social exchange theory is slippery because it exists in such diverse forms, but cognitive heuristics too are not just psychological reductionism but also can be culturally laden.

22Some of the other theoretical perspectives uncovered in the section on theory-in-use can be categorized in terms of decision-making and strength of cultural factors. Interviewer-focused approaches, examined by contemporary researchers such as Morton-Williams (1993), Groves and Couper (1998), de Leeuw (1999) and Lemay and Durand (2002) along with the classic work by Hyman (1954), cover the whole middle row of Figure 2 because the approaches are so varied. Along with the often less successful script (Morton-Williams, 1993:82), in which carefully prepared words are used as heuristic hooks, is participation on face-to-face surveys as "a complex decision process" in which the most successful interviewers manoeuvre and negotiate (Morton-Williams, 1993:103).

23General attitude toward the survey method as a factor guiding specific decisions, another theme from our content analysis of current theory usage, seems to belong in the centre of Figure 2. Attitude toward the survey does not call for exceptionally low or high decision making in that attitudes, a generalizing from experiences and a cross-over concept between psychology and sociology, form gradually.

24As Groves and Couper (1998: 32) argued, the influence of socio-demographics is partly through intervening variables such as attitude toward surveys. That does not make SES, for example, unimportant, but rather suggests an "interpretation" (Kendall and Lazarsfeld, 1950:157). A socio-demographic such as SES affects survey participation on a variety of fronts, not just attitudinally but in terms of the calculation of self-interest, likelihood of deference to authority (Groves and Couper, 1998: 126, 33), and issues of self-disqualification (Morton-Williams, 1993:103; Brehm, 1993:63). In a sense, socio-demographics partition societies into subsets pointing to possible interaction effects in the style of leverage-salience analyses (Groves et al., 2000; Groves et al. 2004).

25Our suggestion, in sum, is that rather than letting different theories talk past each other or having one simply swallow the other undigested, it is better to speculate about the underlying dimensions of theories of survey response behaviour. The ideas outlined here amount to what Turner (1991:9-10, 21) calls a "sensitizing scheme." It is deliberately abstract and covers a wide scope. As Stoop (2005:101) suggested, "social exchange theory is more a model than a theory that can be refuted."

Interpretations and Implications

26Examining the relationship between social exchange theory and cognitive heuristic theory gives a reminder of how thoroughly multidimensional are the reasons why people do or do not participate when asked to do a survey. That is by no means an amazing news flash for most survey methodologists, but even so we found the range of theoretical perspectives invoked impressive, when reviewing the past three years of journal articles on nonresponse. Methodologists believe that people respond to surveys after concluding that the rewards outweigh the costs; that a structure of social disorganization may impede the endeavours of survey researchers; that some socio-demographic pockets of a population are somewhat predestined to refuse or answer a survey; that in some survey settings the decision is made nearly instantaneously, based on heuristic clues; that there are cognitive aspects around how people process information; that the culture is especially sensitized to intrusions on people's time and privacy; that some people have attitudes about surveys in general, apart from topic; that part of the survey response story is physiological; that the interviewer is key, on non-self-administered surveys; that the salience of topic has huge variance, with consequent implications and interactions with other factors around response decisions; that participation is partly an emotional decision; that many obstacles are removed if a surveyed group takes ownership of the survey; that people divide on their amenability to one mode of contact over another; and that the baseline for any theory of response posits these decisions as simply random acts.

27Even though some of these ideas do not always work out as expected (e.g., Hox et al., 1995), this plurality or multi-dimensionality of theorizing deserves emphasis because it gives a clear answer to the two leading research issues of the day around survey response behaviour. First, despite all the improvements in fieldwork technique, interviewer training, deployment of effective cognitive heuristics, why have response rates on most kinds of survey fallen rather than risen over the past decade or two? Answer, so many other factors have worked against the survey researcher, including possibly declining attitudinal support for the whole method, a sense of intrusion arising from saturation surveying (itself part of a larger syndrome as per Wurman, 1989), a possible long term social disorganization in the form of retreatism and disengagement as suggested by Putnam (2000), and a rise in the proportion of people with debilitating ailments such as clinical depression. The handling of declining survey response rates is a little analogous to a city which discovers its crime rate to be rising. The first instinct is to put more police officers on the streets (=improve field techniques); solutions to remove the conditions conducive to crime come later and are less clear (=increasing public awareness of the benefits of good survey research).

28Second, consider the doubt arising in recent years among survey researchers as to whether nonresponse is in fact fairly "ignorable" in terms of biasing results from surveys, an issue aired at the 2005 AAPOR Conference (Groves, 2005). How is it that once the academic community discounted any survey of less than around 50% response rate (the familiar guideline from the Babbie text [1973:165]) yet Keeter et al. (2000) found that a survey of 36% response performed about as well as the same field design augmented up to 61% by extra effort? Answer: it is the multidimensionality of reasons for survey response decisions that lets one nonresponse error cancel out another. Within any single survey setting, assuming salience, mode, sponsor etc. constant, one person may make decisions quickly, heuristically, and move on, another is an amateur philosopher who obsessively takes every decision back to first principles. One is inner-directed not very affected by the cultural climate for or against surveying, another is the other-directed leaf on the cultural breezes. It is this very heterogeneity and diffuseness in theorizing about survey participation that may, due to the cancelling of errors paradox, save the survey method in the 21st century.

Haut de page

Bibliographie

Babbie, Earl R. (1973) Survey Research Methods. Belmont, Calif.: Wadsworth.

Beach, Lee Roy, and Terence R. Mitchell (1978) "A contingency model for the selection of decision strategies", Academy of Management Review 3:439-449.

Befu, Harumi (1977) "Social Exchange", Annual Review of Anthropology 6:255-281.

Berry, Sandra H., and David E. Kanouse. (1987) "Physician response to a mailed survey: an experiment in timing of payment", Public Opinion Quarterly 51:102-114.

Blau, Peter M. (1963) The Dynamics of Bureaucracy. Second Edition. Chicago: University of Chicago Press.

Blau Peter. M. (1964) Exchange and Power in Social Life. New York: Wiley.

Brehm, John (1993) The Phantom Respondents: Opinion Surveys and Political Representation. Ann Arbor: The University of Michigan Press.

Camic, Charles (1986) "The matter of habit", American Journal of Sociology 91:1039-1087.

Cialdini, Robert B. (1993) Influence: The Psychology of Persuasion. New York: William Morrow.

Cox, E.P. (1976) "A cost/benefit view of prepaid monetary incentives in mail questionnaires", Public Opinion Quarterly 40:101-4.

de Leeuw, Edith (1999) "How do successful and less successful interviewers differ in tactics for combating survey nonresponse?", Bulletin de Méthodologique Sociologique 62: 29-42.

de Leeuw, Edith, and Wim de Heer (2002) "Trends in household survey nonresponse: a longitudinal and international comparison", In Survey Nonresponse, editors Robert M. Groves, Don A. Dillman, John L. Eltinge and Roderick J.A. Little, pp. 41-54. New York: Wiley.

Dillman, Don A. (1978) Mail and Telephone Surveys: The Total Design Method for Surveys. New York: Wiley.

Dillman, Don A. (2000) Mail and Internet Surveys: The Tailored Design Method. Second Edition. New York: Wiley.

Dillman, Don A., Joye J. Dillman, and Carole J. Makela (1984) "The importance of adhering to details of the Total Design Method (TDM) for mail surveys", in Source Book 22, New Directions for Program Evaluation, editor Daniel C. Lockhart, pp. 49-64. San Francisco: JosseyBass.

Durand, Claire, A. Blais, and M. Lemay (2002) "Cultural aspects of survey non-response, an empirical assessment", in Social Science Methodology in the New Millennium. Proceedings of the 5th International Conference on Logic and Methodology, editors Jörg Blasius, Joop Hox, Edith de Leeuw, and Peter Schmidt. Verlag Leske and Budrich, ISBN 3-8100-3316-2.

Emmerson, R.M. (1976) "Social exchange theory", in Annual Review of Sociology, editors Alex Inkeles, James Coleman, and Neil Smelser, pp. 335-62. Palo Alto: Annual Reviews, Inc.

Gallegos, Jean G. (1974) An Experiment in Maximizing Response to Telephone Interviews Through Use of a Preliminary Letter, Based on the Principles of Exchange Theory. Master's Thesis, Washington State University.

Gergen, Kenneth J., Martin S. Greenberg and Richard H. Willis (1980) Social Exchange: Advances in Theory and Research. New York: Plenum.

Gordon, W. Terrence (1997) Marshall McLuhan: Escape into Understanding. Toronto: Stoddart.

Gouldner, Alvin (1960) "The norm of reciprocity: a preliminary statement", American Sociological Review 25:161-78.

Goyder, John (1987) The Silent Minority: Nonrespondents on Sample Surveys. Cambridge: Polity Press.

Groves, Robert M. (1989) Survey Errors and Survey Costs. New York: Wiley.

Groves, Robert M. (2005) "Total Survey Error", part of a plenary session at the 60th Annual AAPOR Conference, Miami, May 12-15.

Groves, R.M., R.B. Cialdini, and M.P. Couper (1992) "Understanding the decision to participate in a survey", Public Opinion Quarterly 56, 475-95.

Groves, Robert M., and Mick P. Couper (1998) Nonresponse in Household Interview Surveys. New York: Wiley-Interscience.

Groves, Robert M., Don A. Dillman, John L. Eltinge and Roderick J.A. Little (2002) Survey Nonresponse. New York: Wiley Inter-science.

Groves, R., E. Singer and A. Corning (2000) "Leverage-salience theory of survey participation: Description and an illustration", Public Opinion Quarterly 64: 299-308.

Groves, R, S. Presser and S. Dipko (2004) "The Role of Topic Interest in Survey Participation Decisions", Public Opinion Quarterly 68: 2-31.

Gunn, Walter J., and Isabelle N. Rhodes (1981) "Physician response rates to a telephone survey: effects of monetary incentive level", Public Opinion Quarterly 45:109-115.

Halpern, S.D., P.A. Ubel, J.A. Berlin, and D.A. Asch (2002) "Randomized trial of 5 dollars versus 10 dollars monetary incentives, envelope size, and candy to increase physician response rates to mailed questionnaires", Medical Care 40:834-839.

Hauser, Robert M. (2005) "Survey response in the long run: the Wisconsin longitudinal study", Field Methods 17: 3-29.

Hawkins, D.F. (1975) "Estimation of nonresponse bias", Sociological Methods and Research 3:461-88.

Heath, Anthony (1976) Rational Choice and Social Exchange: A Critique of Exchange Theory. Cambridge: Cambridge University Press.

Homans, George C. (1958) "Social behavior as exchange", American Journal of Sociology 63:597-606.

Hox, J., E. de Leeuw, and H. Vorst (1995) "Survey participation as reasoned action: a behavioural paradigm for survey nonresponse?", Bulletin de Méthodologie Sociologique 47:52-67.

Hyman, Herbert (1954) Interviewing in Social Research. Chicago: University of Chicago Press.

Johnson, Timothy P., D. O'Rourke, J. Burris, and L. Owens (2002). "Culture and survey nonresponse", in Survey Nonresponse, editors Robert M. Groves, Don A. Dillman, John L. Eltinge and Roderick J.A. Little, pp. 55-69. New York: Wiley.

Kaplowitz, M.D., and F. Lupi (2004). "Color photographs and mail survey response rates", International Journal of Public Opinion Research 16:199-206.

Keeter, S., A. Kohut, A. Miller, R. Groves, and S. Presser (2000) "Consequences of reducing nonresponse in a large national telephone survey", Public Opinion Quarterly 64:125-148.

Kendall, Patricia L., and Paul F. Lazarsfeld (1950) "Problems of survey analysis", in Continuities in Social Research: Studies in the Scope and Method of "The American Soldier," editors Robert K. Merton and Paul F. Lazarsfeld, pp. 133-196. Glencoe: Free Press.

Kuper, Adam (1983) Anthropology and Anthropologists: The Modern British School. Revised edition. London: Routledge and Kegan Paul.

Lemay, Michael, and Claire Durand (2002) "The effect of interviewer attitude on survey cooperation", Bulletin de Méthodologie Sociologique 76: 26-44.

Linsky, A.S. (1975) "Stimulating responses to mailed questionnaires: a review", Public Opinion Quarterly 39:82-101.

Makoba, Wagona J. (1993) "Toward a General Theory of Social Exchange", Social Behaviour and Personality 21:227-240

Malinowski, Bronislaw (1926) Crime and Custom in Savage Society. London: Routledge and Kegan Paul.

Menzies, Ken (1982) Sociological Theory in Use. London: Routledge and Kegan Paul.

Mizes, J. Scott, E. Louis Fleece, and Cindy Roos (1984) "Incentives for increasing return rates: magnitude levels, response bias, and format", Public Opinion Quarterly 48:794-800.

Morton-Williams, Jean (1993) Interviewer Approaches. Aldershot, U.K.: Dartmouth.

Putnam, Robert D. (2000) Bowling Alone: The Collapse and Revival of American Community. New York: Simon and Schuster.

Riesman, David (1950) The Lonely Crowd: A Study in the Changing American Character. New Haven: Yale University Press.

Rizzo, L., J.M. Brick and I. Park (2004) "A minimally intrusive method for sampling persons in random digit dial surveys", Public Opinion Quarterly 68: 267-74.

Stoop, A.L. (2005) The Hunt for the Last Respondent: Nonresponse on Sample Surveys. The Hague: Social and Cultural Planning Office.

Stock‚, V., and B. Langfeldt (2004) "Effects of survey experience on respondents' attitudes towards surveys", Bulletin de Méthodologie Sociologique 81:5-32.

Teitler, J.O., N.E. Reichman and S. Sprachman (2003) "Costs and benefits of improving response rates for a hard-to-reach population", Public Opinion Quarterly 67: 126-38.

Turner, Jonathan H. (1991) The Structure of Sociological Theory. Fifth edition. Belmont, California: Wadsworth.

Warriner, Keith, John Goyder, Heidi Gjertsen, Paula Hohner, and Kathleen McSpurren (1996) "Charities, no; lotteries, no; cash, yes: main effects and interactions in a Canadian incentives experiment", Public Opinion Quarterly 60:542-62.

Wrong, Dennis H. (1961) "The over-socialized conception of man", American Sociological Review 26:183-93.

Wurman, Richard Saul (1989) Information Anxiety. New York : Doubleday.

Zafirovski, M. (2003) Some Amendments to Social Exchange Theory: A Sociological perspective. CAAP. Retrieved on March 22, 2004 from http://theoryandscience.icaap.org/content/vol004.002/01_zafirovski.html

Zafirovski, M. (2005) Social Exchange Theory under Scrutiny: A Positive Critique of its Economic-Behaviorist Formulations. Electronic Journal of Sociology. Retrieved on December 10th 2005 from www.sociology.org/content/2005/tier2/SETheory.pdf.

Haut de page

Notes

1  Throughout this discussion we are setting aside the issue of achieving contact with sample listings.

2  It cannot be claimed that this coding is an entirely scientific, objective exercise that another observer would reproduce 100%. Nevertheless, we worked on the codes over a period of a year and a half so that first impressions could be re-visited many months later. Decisions made earlier on the whole still appeared reasonable as much as a year later. Reference to key authors such as Dillman (2000) or Groves and Couper (1998) were considered meaningful for clarifying hints at social exchange theory and cognitive heuristic theory respectively.

3  As Groves and Couper (1998: 130) observed, exchange theory is very "diffuse," but also their data were too diffuse: information was needed about the attitude of the sample listings, both respondents and nonrespondents, high SES and low SES, toward the US government. Without that, any hypothesizing was shooting in the dark.

Haut de page

Pour citer cet article

Référence papier

John Goyder, Luc Boyer et Guil Martinelli, « Integrating Exchange and Heuristic Theories of Survey Nonresponse* », Bulletin de méthodologie sociologique, 92 | 2006, 28-44.

Référence électronique

John Goyder, Luc Boyer et Guil Martinelli, « Integrating Exchange and Heuristic Theories of Survey Nonresponse* », Bulletin de méthodologie sociologique [En ligne], 92 | 2006, mis en ligne le 01 octobre 2009, consulté le 23 octobre 2019. URL : http://journals.openedition.org/bms/554

Haut de page

Auteurs

John Goyder

Department of Sociology, University of Waterloo; jgoyder@uwaterloo.ca)

Luc Boyer

Department of Sociology, University of Waterloo

Guil Martinelli

College of Nurses of Ontario

Haut de page

Droits d’auteur

© BMS

Haut de page
  • OpenEdition Journals