Steps Toward an Information Based Theory of Attitudes

Davis Foulger
May 4, 1979
International Communication Association
Philadelphia, PA

(more papers by Davis Foulger)


Steps are taken in the direction of constructing a theory of attitude formation and change based in the concepts of information processing and the variables of distraction. A theoretical vocabulary, based in the primitive terms information and evaluation, is derived. The concept of attitude critically examined with reference to these primitives, the concept of information processing, and the notion of cueing. The resulting conception of attitude is sensitive to variations within the situation that can affect attitudes. An attempt is then made to construct an algebra within which attitude variables could be expressed. Although the theory, as presented, cannot be considered complete, it is believed that sizable steps are taken in the direction of an information processing based theory of attitude formation and change.

Primitive Terms

Theories of attitude have generally constructed attitudes out of clusters of beliefs. (Festinger, 1957; McGuire, 1964; Rokeach, 1968; Fishbein and Ajzen, 1975). Although attitude is best treated as a derived term within a theory of attitude, treatment of belief as a root term to such a theory leaves something to be desired, especially when attitudes are examined from the perspective of information processing. This paper attempts to make just such an examination, integrating literature from human information processing (Norman, 1968; Mazzaro, 195), general theories of attitude (Festinger, 1957; McGuire, 1964; Rokeach, 1968; Fishbein and Ajzen, 1975; Wyer, 1974, and Cappella and Folger, 1977), and research concerning the effect of distraction on persuasion (Festinger and Maccoby, 1964; Osterhouse and Brock, 1968; Baron, Baron, and Miller, 1973; Foulger, 1977 and others).

The very idea of examining attitudes from the perspective of human information processing suggests that information is a root term for attitude theories. If attitudes do reflect our experience of the world, it is reasonable to expect that it is from that experience, from the information we have concerning an attitude object, that we construct attitudes. Even if this were not true, moreover (and it is hard to imagine it not being), the attitude itself is information.

Evaluation is proposed as a second root term to an attitude theory. The word, unlike information, is less than ideal as a primitive in that it can be defined in terms of information. It does not appear, at this point, that any economy would be accomplished through such definition. Evaluation attempts to tap that part of an attitude which is a judgment or evaluation. An attitude is not merely a piece of information; it is a piece of evaluative information, information concerning an auditors evaluation of an attitude object.

It is hoped that information and evaluation will provide a sufficient core of primitives for a theory of attitude. As primitive terms to this presentation, they will be used to define all of its key terms.

Basic Derived Terms

The Structure of Memory

Memory is now commonly depicted as having a network format. Arguments for such a format can be found in Cappella and Folger (1977, from the perspective of attitude) and Lindsey and Norman (1977). The network model represents memory as an n-dimensional collection of concepts and events, depicted as nodes, which are interconnected by their temporal order, spatial relationship, and/or concept al relationship. The nature of these connections, which are really associations between concepts and events, is covered in Lindsey and Norman (1977, c. 10).

Two forms of network memory have been supported and it has been suggested that both exist. The first, episodic memory, retains information in a way analogous to the way it was experienced, including both concepts (people, places, objects, etc.) and events (the actions of people, objects, etc.). These objects and events are felt to be remembered in the form concept--event--concept, analogous to the sentence form subject--verb--object. The second form, semantic memory, is memory of relationships experienced among concepts. In semantic memory, the more situationally specific connections of episodic memory are collapsed to more general concept--concept links. If both episodic and semantic memory do exist, it may well be that semantic memory is derived from episodic memory. This derivation is partially demonstrated by a linguistic analogy in which the sentence form 'subject/(verb/object)' (concept--event--concept) reduces to 'noun phrase/verb phrase' (concept--concept). The existence of both has interesting implications for (and may gain partial support through) attitude measurement. This will be explored below.

The Nature of Beliefs

Whether one follows Rokeach's (1968) definition (a belief is a statement that can be prefaced with the words 'I believe that...') or Fishbein and Ajzen's (1975) definition (a belief is information a person has about an object; information that specifically links an object and attribute), a belief is a sequence of connected information which an individual 'believes in'. Cappella and Folger (1977) locate two forms which beliefs can take. The first, the concept--concept belief, corresponds to Fishbein and Ajzens's definition and the structure of connections in semantic memory. The second, concept--event--concept beliefs, correspond more closely to Rokeach's definition and to the structure of connections in episodic memory. As support exists for for the existence and use of both types of beliefs and reason exists for believing that both corresponding forms of memory exist, it is reasonable to infer that both forms of belief do occur.

Beliefs must be more than the connection of information in memory if they are to be useful for constructing attitudes. As attitudes are evaluations, beliefs must involve some form of evaluation, some form of affective connection, for attitudes to be formed from them. It is not beliefs, but the affective connections to beliefs, which are clustered to form attitudes. Thus it is not wholly true that beliefs are clustered to form attitudes. Beliefs must, however, be accessed if the evaluations connected to those beliefs are to be clustered, so it is not wholly untrue either.

Attitude measurement measures the affective connections of beliefs. Interestingly, it can do so at least three different ways. The first approach, the semantic (concept--concept) approach, subsumes most current measures, including paired comparison, Likert scaling, semantic differential, and related attitude scaling techniques. This approach generally uses multidimensional or linear weighting schemes to combine a series of concept--concept comparisons, one measure for each belief. The key feature of such measures is the existence of reference points against which belief statements or concepts can be judged.

There may be as few as one such points (self -- Saltiel and Woelful, 1975) or as many as are contained within a semantic differential network. The effect of such measurement is to define a network of affective connections to beliefs or concepts around an attitude object and describe attitudes in those terms. Unless all of the important elements of this network are specified (as in the case of Fishbein and Ajzen's (1975) intentions) the attitude measured will be a compromise of several possible attitudes. This occurs because belief concepts used are collapsed from the many situationally specific relationships of episodic memory.

The second approach, the episodic (concept--event--concept) method, obtains evaluations of each concept and event which occurs in any given belief statement and combines these evaluations and their product terms to a weighted linear equation. The equation yields an evaluation of the statement, an attitude toward the statement which, because of the episodic nature of the information, should be highly situation specific. The approach, in a sense, turns the semantic approach inside out in that the episodic method often uses evaluations of concepts to evaluate statements while the semantic method often uses evaluations of statements to evaluate concepts. The relationship may be better stated, however, as one of specificity, with episodic measures specifying explicitly measures which semantic measures may only tap implicitly. Measurement of episodic evaluations is difficult for the researcher (See Cappella and Folger, 1977, for a review of the method) and, very likely, for the individual. This should be especially true when the individual must face several such evaluations in the course of evaluating an attitude object.

Although the brain is capable of wondrously complicated feats, the complexity of episodic evaluations suggests that they are probably not subject to reprocessing but are made once and thereafter attached to the episodic memory of which they are part. An affective evaluation of this sort, attached to each belief in episodic memory, would greatly simplify the retrieval of affective information from episodic memory.

A third approach to attitude formation builds on this simplification, forming attitudes out of an individual's evaluations of individual beliefs. In straightforward terms, this approach sums evaluations of relevant beliefs evaluated either semantically or episodically. Relevant beliefs can be sorted, according to their affective connections, into message counter beliefs and message consonant beliefs to yield the following statement of how attitudes are formed in this third approach:


Equation 1

where A is the attitude toward an object, c and d are, respectively, message consonant and counter beliefs, and e are evaluations of those consonant and counter beliefs.

It should be noted that beliefs can also be classified according to their acceptance or rejection by the individual. Rejected beliefs should play little role in attitude formation except to provide reference points for the speedy rejection of previously rejected beliefs before they become an active focus of attention. The existence of such a text of rejected beliefs makes sense. Sufficient processing should have occurred in their initial rejection to commit them to memory. Their existence and functioning as above would, moreover, go a long way toward explaining selective perception (biased scanning). Accepted beliefs should, of course, be used in attitude formation.

The Nature of Attitudes

A central characteristic of definitions of attitude is a statement of their enduring nature, a statement which implies that since attitudes change little, that the processing of attitude information is minimal in the day to day usage of formed attitudes. Studies have supported the idea of the attitude as enduring over months, years, and even decades, given a reinforcing environment, but it remains difficult to to account for either the variety and malleability of attitudes or their inconsistency in predicting behavior without appeal to some sort of quasi continuous processing and reprocessing of information and attitudes. If attitudes are reformed periodically on the basis of information, the question becomes one of how and what information is accessed and combined to form attitudes. The question of accessing is attacked by Cappella and Folger (1977) with cueing.

Cues are generated through the information we receive from messages, communicators, and the context (situation) within which the message occurs. A cue directly accesses information related directly to the cue and indirectly accesses additional information connected to the directly cued information. By bringing information into the focus of attention, cueing provides access to beliefs which can then be used in attitude formation, counter argument with a persuasive message, or simply thinking.

Cued beliefs concerning an attitude object are arguments. The affective connections to cued beliefs argue, following equation one, for specific attitudes. Attitudes are formed, in turn, on the basis of these arguments. Beliefs which are not cued do not, of course, affect the formation of those attitudes. If an individuals resources for processing information are, as researchers in human information processing claim, limited, then cueing makes sense as a mechanism for selecting a manageable subset from a large cluster of beliefs concerning an attitude object. This selection not only allows for variation in attitudes as cues change (as say in an experiment) but for the generation of highly context specific attitudes tailored to the specific situation.

This enables the offering of the following definitions:

An Introduction to Distraction

From the standpoint of influence, a distraction is a second message presented to an auditor/audience at the same time that a first persuasive message is being presented to that same auditor/audience. The effect of this second message, as first hypothesized by Festinger and Maccoby (1964), is to increase the persuasive effect of the persuasive message with which it is presented by distracting the auditor from arguing against the persuasive message (counter arguing). Explanation of the effect of distraction has not proved simple. A variety of explanations have been proposed to explain the effect. Many have been nourished by conflicting results obtained in the course of research. For a review of these proposals, the reader is referred to Baron, Baron, and Miller (1973) and Foulger (1977). The present view is that the correct explanation lies in the joining of two of these explanations (Festinger and Maccoby, 1964; and McGuire, 1966) stated in terms of information processing. The work of Wyer (1974) and Cappella and Folger (1977) is particularly useful in this synthesis.

The process of generating an information processing explanation for distraction is aided by the existence of a parallel notion, one of 'interference' within cognitive information processing research. The various forms of interference explored within this notion is more diverse than will be of interest here, but a large class of interfering stimuli involve the reception of simultaneous messages both within and across modalities, with the detection, reception, and memorization of information contained within two separate messages which are addressed to the subject at the same time. Although this research operates at a very different level than that approached in distraction research, the parallels between the two approaches should make examination of distraction from the perspective of interference an interesting and fruitful enterprise.

Festinger and Maccoby's (1964) explanation of the effect of distraction on persuasion centered around counter argument reduction, the view that the distracting message drew the attention of the auditors away from refuting and countering the (message consonant) arguments offered by the persuader. This explanation makes the implicit assumption that attitudes are formed from those ideas (beliefs, arguments) which we remember and hold to be important. McGuire (1966) suggested that the effect of distraction should be radically different than the effect proposed by Festinger and Maccoby (1964); that the distraction should draw attention away from listening to (and comprehending) the persuasive message. The resulting reduction in comprehension, it was argued, should result in a decrease in the persuasive effect of a message.

Festinger and Maccoby's (1964) distraction hypothesis assumes a particular model of attitude formation which states that attitudes are formed in the comparison of message counter and consonant arguments or beliefs. The model was explicitly stated above as equation one above. This model is not uncommon in the literature and is assumed by both Festinger (1957) in the theory of cognitive dissonance and McGuire (1964) in inoculation theory. The implication of Equation 1 is that attitude can be changed by either increasing or decreasing either message consonant or message counter argument.

A wide variety of manipulations testing the distraction hypothesis have been constructed since the original Festinger and Maccoby manipulation. Although: many of these manipulations have supported both distractions effect on persuasibility and the counter argument disruption explanation, manipulations are not always effective in supporting either hypothesis. Although McGuire's (1966) learning hypothesis is no more effective in explaining the bulk of distraction results (It actually fares considerably worse), where results cannot be explained by counter argument reduction, there is often evidence of reduced comprehension. Even with this consistency in the results, however, distraction results are anything but' simple to explain. At least six different variables have emerged which have some impact on the result of a distraction manipulation. It is valuable to remember, in examining these variables, that the distraction itself is a variable.

Distraction: An Information Processing Explanation

A distraction is a message. Both it and the persuasive message with which it occurs offer information to the auditor. Comprehension of either message requires that an auditor attend those messages. Researchers in the area of human information processing have used research strategy not unlike that used in distraction research in order to explore the nature of attention. This research, which is reviewed in depth in Mazzaro (1975, c. 12 - 16) and Norman (1976) has revealed a great deal about how attention affects the detection, recognition, and rememberence of information. This research amply supports the following three points:

P1: Our ability to attend and process information is rather flexible.

P2: There is a limit to our ability to Process information .

P3: Even unattended information is processed to a minimal level, a level sufficient to interfere with our processing of attended information.

Attention, as it is presented in the information processing literature, is a process of making connections between incoming "bottom up" (Norman, 1976) information and the "top down" knowledge or "text" (Foulger, 1978) already maintained in our memory. Attention operates, in the sense portrayed here, as a sort of switchboard operator, connecting incoming information with relevant portions of memory and maintaining those connections with memory already established until other connections take precedence. An auditor faced with attending and comprehending both a persuasive message and a distracting message is faced with a competition for limited processing resources. Although the auditor has reasonable flexibility in how she or he attends the competing messages, there is a limit on the amount of information which will be processed. If this limit is exceeded, comprehension of one or both of the messages will suffer. If, moreover, the auditor, realizing this, undertakes to ignore the distraction, it will still be processed to a minimal level sufficient to interfere with the processing of the persuasive message.

If the above analysis told the whole story, one would embrace McGuire's (1966) suggestion without question, but the distraction and the persuasive message are only two of at least three messages which the auditor is processing in the distraction experiment. The auditor also generates a message in which he or she recalls from memory arguments (information) which support or counter the arguments presented in the persuasive message. The co-occurrence of these three messages - influence, distraction, and counter-influence - should at least strain the limits of the auditors processing resources. It is here that the Festinger and Maccoby (1964) and McGuire hypotheses join. As long as the distracting message is attended as being less important than the persuasive message, the effect of the distraction should be modest, one of reducing counter argument (attention to the third message) and increasing persuasion as predicted by Festinger and Maccoby (1964). But as soon as the distraction becomes the primary message, its effects become far more severe, with comprehension of the persuasive message declining and a reduction in its persuasive effect. Both effects are, although quite different in terms of their effect on influence, quite similar insofar as both involve a reduction in the processing of messages.

Belief as Information: A Hierarchy

It has been argued at several points in this paper that attention is not committed to all available information. The information selected is taken from a range of available information; the information connections established taken from a range of potential connections. These potential connections exist both in the communication environment and within our thinking. Connections are selected from among the range of potential connections through attention. These connections exist both between incoming information and memory (received connections) and between separate clusters of information within memory (thought). These connections, in turn, can be classified as consonant, counter, and irrelevant with respect to an attitude object. Among received connections, they may derive from primary (persuasive in a theory of attitude) and secondary (distracting, in a theory of distraction) messages. Finally, the connections may or may not involve rejected arguments.

Table 1: A Hierarchy of Information

These information elements can be represented in a pyramid form which shows the relationship among the various elements. This is depicted in figure one, in which PC represents potential connections, C represents connections, C' represents connections not made, T represents thinking, R represents received connections, pm represents primary message connections, sm represents secondary message connections, c represents consonant connections, d represents counter connections, i represents irrelevant connections, represents rejected connections, and r represents accepted connections.

The primary usefulness of this structure is heuristic and explanatory. Measuring potential connections is not possible with current technology. The same is true of cued connections. It may not be possible, moreover to set an upper limit to the number of connections which can be engaged by attention at any given time. The representation usefully viewed as a hierarchy of subscripts which can be usefully substituted into the equations which will be presented in the following sections.

Distraction: Key Variables

The relationship created in this joining assumes the existence of the first of eight distraction variables, that of distraction strength. At low to moderate strengths, distraction attenuates counter arguments and accentuates persuasion. At high strength levels, however, comprehension is attenuated and persuasion suffers. This relationship is described by the inverted-U of figure 2 and in an equation offered by Wyer (1974). Wyer's equation is:

Equation 2

where the P of I is the probability of influence, the P of R is the probability of comprehension, and P of CA is the probability of counter argument. A table for constructing curves with Wyer's equation is shown in table one. Figure two illustrates one possible curve (see Wyer, 1974 for other possibilities). The figure was constructed by moving down a column (increases in distraction strength first decrease counter argument with comprehension held constant) and then moving across a row (comprehension decreases and counter argument remains constant as distraction strength reaches high levels). Influence will proceed from low to high to low.

Figure 2: The inverted U-shaped curve representing the effect of strength of distraction on influence.


Evidence for the existence of a varying distraction strengths is available in a number of experiments, including Rosenblatt (1966), Osterhouse and Brock (1970), Keating and Brock (1974), Zimbardo, Thomas, Snyder, Gold, and Gewertz (1970), and can be strongly inferred by the diverse results of manipulations like those of Allyn and Festinger (1971), Festinger and Maccoby (1964), Freidman and Sears (1965), Haaland and Venkatesan (1968) and Vohs and Garrett (1968). The evidence indicates that distraction strength can be broadly represented by a typology. The weakest strengths of distraction are self-induced. This level of distraction is experienced when the auditor creates a second message by concentrating on something other than the persuasive message (the personality of the speaker, a good looking woman or man, etc.). Research on this form of distraction (Freidman and Sears, 1965) indicates that the effect of such distraction can be quite weak, but personal experience of daydreaming should provide, for those who have experienced or observed the phenomena, strong phenomenological and/or ethnographic evidence of the potential strength of self- induced distraction. This category of distraction can also be examined as a variable (perceptual focus) which can affect other forms of distraction.

Table 1: The probability of influence as a function of probability of counter argument. See equation 2. From Wyer (1974, p. 190.

Moderate levels of distraction correspond to environmental distractions, actual second message distractions. These are distractions (irrelevant film or slides) which are interesting enough to attract the attention of the subject, but which do not require the active participation of the subject. In life, they occur but are beyond our control. They act to reduce counter argument only under all but the strongest conditions, those in which they demand recognition as the primary message. The third category, requires the auditor to engage in some form of problem solving while listening to a persuasive message and might be called evaluative. This type of distraction forces the auditors attention to the distraction, making it the primary message. It should hardly be surprising that comprehension of the persuasive message declines under these circumstances.

Relevance, the extent to which two or more messages act as a single message, is a second variable which affects the way in which a distraction affects a persuasive message. When a distraction is irrelevant to the persuasive message, the effect of the distraction is that of competition for attention, decreasing counter argument, and, where the distraction is strong, decreasing comprehension. Where the distraction is relevant to the content of the persuasive message, its effect can be expected to differ substantially, with the distraction being actually integrated with the persuasive message. Instead of drawing attention away from the message, the distraction draws attention to the message. These messages may include a speakers visual aids and the reaction of audience members to the message and will have the force of either directly reinforcing or denigrating the arguments used by the speaker. Research in the area of is limited, so far as the author is aware, to two studies of heckling (Ware and Tucker, 1974; Silverthorne and Mazmanian, 1975) which hypothesize that the heckling will distract and increase influence but find decreased influence.

A third variable, the evaluative bias of the relevant distraction, can be inferred. Positive distractions, those that re-enforce the persuasive message, should be expected to increase persuasive effect of the message. Negative distractions, those that denigrate the content of a persuasive message, should be expected to decrease the persuasive effect of the message by pointing out message counter arguments.

Distractions which are relevant to a persuasive message affect an auditor of_that persuasive message in ways that are very different from the effect generated from a relevant distraction. The information contained within the relevant distraction still competes for processing resources, but it competes for linking to the same memories cued by the persuasive message. Thus the relevant distraction can be said to build on the persuasive message. The irrelevant distraction directly competes for attention, however, not only in terms of processing resources, but in terms of cueing very different ideas. The negative relevant distraction may well attenuate auditor generated arguments which counter the message, but it makes up for this by cueing a set of distraction generated counter arguments. The positive relevant distraction may also attenuate auditor generated argument, but its most important effect is to generate a set of arguments, cued in the distraction, which support the message. The effect of irrelevant distractions is that, and only that, of attenuating auditor generated counter argument and comprehension. Thus distraction strength is not expected to have much impact on relevant distractions even though it is the key variable for irrelevant distractions. While there are undoubtedly many other variables which affect the impact of relevant distractions, the evidence is thin. The variables which remain to be examined here all have their primary effect with irrelevant distractions.

The perceptual focus, the message which is engaging the greatest attention, is the most important of these. Since distractions act as a competing second message to a persuasive message, it is possible to increase and decrease its effect by instructing an auditor to concentrate on either the distraction or the persuasive message, respectively. Focusing the auditor's perceptions on the distraction will increase its effect on influence. Focus on the persuasive message decreases the distraction's effect, although never entirely. The effect of perceptual focus is depicted in figure three. Less distraction strength is necessary when the auditor focuses on the distraction necessary to obtain a given effect when the auditor focuses on the distraction. More is necessary when the auditor focuses on the persuasive effect. Evidence supporting the effect of perceptual focus is found in Zimbardo et. al. (1970) and other studies.


Figure three: The effect of perceptual focus on distraction strength and influence.

It should be remembered that perceptual focus is related to the self-induced distraction. Taken from this perspective, it becomes one of the most interesting distraction variables with the potential to tie into a variety of other persuasion variables, notably source credibility and emotional appeals. Focusing ones attention on the good qualities of the speaker, rather than the speakers message, may have a distracting effect. Similarly, the effect of an emotional appeal may be to focus the listener's attention on his or her feelings rather than on counter arguments, another case of distraction. Wyer (1974) also makes a case for fear appeal being implemented through a process of distraction.

A fifth variable is found in message complexity, the extent of interconnection between the information contained in a message. The effect of this variable is seen in figure four_as intensifying the effect of distractions. This makes sense from the standpoint of information processing. Since complex messages are more interconnected than simple messages, they should require more attention to make all the connections necessary for comprehension than do simple messages. The complex messages increased attentional requirements make it more vulnerable to the competition of a distraction than would be a simple message and as a result, less distraction strength is necessary to accomplish, in a complex message, a given level of counter argument or comprehension reduction than would be required for a simple message.

Topic familiarity, the extent to which a connection between a specific cue and memory has become automatic (the ease with which such connections are constructed), can be expected to operate in two distinct ways. On the one hand, high topic familiarity, it should be easier (require less attention) to counter argue when faced with highly familiar arguments. As a result, distractions should have less effect as shown in the high topic familiarity curve of figure four. On the other hand, low topic familiarity, counter argument will be difficult because there are few, if any available (This is the case of truisms in inoculation theory (McGuire, 1964)). Distraction cannot, as a result, reduce counter argument to any large degree. The distraction should, however, remain capable of reducing comprehension. The resulting figure four curve drops (due to comprehension reduction) after a plateau (due to lack of counter arguments.

Figure four: The effects of message complexity and topic familiarity on the relationship between distraction and influence.

A seventh variable which can affect the impact of a distraction is channel overlap, the extent to which the distraction and persuasive message are received via the same sense. The eighth variable is message similarity, the level of informational overlap between the persuasive message and the distraction. From information processing theory we take the following propositions:

P4: Less processing is required to separate signals presented on a single channel than is required to separate signals delivered to different channels.

P5: The greater similarity between the irrelevant distraction and the persuasive message, the stronger the distraction.

Both have received copious support in information processing research. Channel overlap may also be supported in distraction research by Regan and Cheng (1973), although they did not explicitly test this result. Message similarity requires an explicit test. The effect of channel overlap should be, however, to reduce the effectiveness of the distraction. The effect of message similarity should be to increase the effectiveness of the distract ion.

Distraction Summary

Eight variables which affect the distraction effect on influence have been identified through the literature of information Processing and distraction and organized in a human information processing model. For convenience, these variables are briefly summarized in Appendix I. Similarly, the vocabulary for attitude theory is summarized in Appendix II. It is now time, having given attitude and distraction a contribution of information processing theory, to see what they can give information processing in return.

A Representation for Attention

In order to represent attention it is first necessary to represent connections within attention. This is done in equation three:

Equation 3

where E represents the ease of connection, F is the familiarity of the connection, IC is the simplicity of the connection, and IS is the dissimilarity of the connection to other connections.

The ease of connection determines the number of connections which can be made within the confines of attention. When the ease of connection is high for all connections, many connections can be made. When the ease of connection is low for any connection, many fewer connections can be made. Familiarity raises the ease of connection. Complexity and similarity to other current connections lowers the ease of connection.

Given ease of connection, it becomes possible to represent attention. This is done in equation four:

Equation 4

where a is the attention given an individual connection, k is a constant representing the number of connections which can be maintained when the ease of connection approaches its lower limit of zero (maximally difficult connection). It may not be possible to establish k, even for an individual, but it is convenient to assume a constant representing the limit of our attentional capacity.

This equation is easily extended to messages (clusters of connections) in the manner of equation five:

Equation 5

The Theory as a Representation of Distraction

The effects of three of the outlined distraction variables are accounted for within the concept of ease of connection. Topic Familiarity is represented in the familiarity of the connection. Message Similarity is represented in the measure of dissimilarity. Message Complexity is represented in the measure of simplicity. These variables affect the amount of attention which is available for counter arguing and comprehension.

Perceptual focus affects the allocation of attention by noting that the subject has a choice of where attention will be allocated. Focus is cannot be subsumed within the equations, but can be examined in terms of the hierarchy. Focusing attention on the persuasive message will reduce the attention which can be allocated to the secondary message and vice versa. These choices will, moreover, effect the results obtained in equations one and two, Focus on the distraction should reduce comprehension and prevent increases in message consonant argument. Concentration on the message should increase comprehension and allow increases in consonant argument. Counter argument should not be affected..

Relevance should also see its effects not so much in attention as in choices within the hierarchy and effects in equations one and two. Relevant distractions should add cues to those presented in the message. Evaluatively positive cues should result in increased consonant argument. Negative cues should result in increased counter argument. Either effect can be traced in equation two.

Channel of reception has no sure explanation within the model. The variable points to the theory's need for further refinement.

Strength of distraction requires some additional theory, represented in equation six:

Equation 6

where SD is the strength of distraction. The measure is a ratio of the number of message irrelevant potential connections to the total potential connections. The larger the ratio, the greater the potential for distraction. This is so simply because greater amounts of message irrelevant material are competing for attention. Once again, the effect of strength of distraction while argued at the level of information and attention, is not directly represented within the attention equation. However, its effect can be traced through the hierarchy and equations one and two.


The goals of this paper, introducing information processing concepts to attitude theory and, more directly, to distraction research, the development of a vocabulary for attitude theory, the development of an information Processing structure for attitude theory and measurement, the identification and organization of key variables in distraction theory and research, the development of a cohesive theory of distraction's effects, and the development of a theoretical calculus to represent the concepts, vocabulary, variables, and theory so developed, has been ambitious, the work of several years and, in the normal course of things, several papers. It is believed that the paper has succeeded, albeit with great brevity, in examining all of the above. The theory is not complete. Far from it. But the existing structure seems to fit together rather well, represent the state of the research rather well, so there is reason to hope that it represents steps in the right direction for understanding attitudes.

A variety of strategies for accomplishing attitude change are suggested within the formulation. First, the addition of new information to memory opens up the possibility that, when cued, that new information will affect attitude in the way desired by the provider of the information. Second, persuading the individual that information already in memory should be rejected should prevent the cueing of that information in future attitude formation. Such rejection should, moreover, provide a basis for biased scanning in future messages. Third, the subject can be persuaded to change the evaluation implied in existing affective connections. This should effect the impact of such affective connections in future cueings. Rokeach (1968) provides a classic example of such affective change. Fourth, the persuader can manipulate cues so as to obtain a desired, if short lived, situational attitude. Fifth, the variables of distraction can be manipulated with the effects implicit to this theory.


The author has become indebted, at various points in the course of this work, to Albert Pryor of The University of Central Florida, C. David Mortensen and Joseph Cappella of the University of Wisconsin, and Herbert Simons and Thomas Steinfatt of Temple University. This paper could not be what it is today without their help and criticism.


Allyn, J. & Festinger, L. The effectiveness of unanticipated persuasive communication. Journal of Abnormal and Social Psychology, 1961, 62, 35-40.

Baron, R. S., Baron, P. H., & Miller, N. The relation between distraction and persuasion. Psychological Bulletin, 1973, 80, 310-323.

Brook, T. C. Communication discrepancy and intent to persuade as determinants of counter-argument production. Journal of Experimental Social Psychology, 1967, 3, 296-309.

Cappella, J. & Folger, J. An Information Processing Explanation of the Attitude Behavior Inconsistency. Unpublished Manuscript, University Of Wisconsin, 1977.

Festinger, L. A Theory of Cognitive Dissonance. Evanston, Ill.: Row, Peterson, 1957.

Festinger, L., & Maccoby, N. On resistance to persuasive communications. Journal of Abnormal and Social Psychology, 1964, 68, 359-366.

Fishbein, M., & Ajzen, I. Belief, attitude, intention, and behavior. Reading, MA: Addison-Wesley, 1975. -

Foulger, D. A. Distraction and Dissonance: A model of the persuasive process. Unpublished Masters Thesis, Florida Technological University, 1977.

Foulger, Davis A. An examination of the concept of text. Unpublished Manuscript.. U. of Wisconsin, 1978.

Friedman, J. L. and Sears, D. O. Warning, distraction, and resistance to influence. Journal of Personality and Social Psychology. 1965, 1, 262-266.

Haaland, G. A, and Venkatesan, M. Resistance to persuasive communications: An examination of the distraction hypothesis. Journal of Personality and Social Psychology, 1968, 9, 167-170.

Keating, J. P., & Brock, T. C. Acceptance of persuasion and the inhibition of counter argumentation under various distraction tasks. Journal of Experimental Social Psychology, 1974, 10, 301-309.

Kiesler, S. B., & Mathog, R. B. The distraction hypothesis in attitude change: The effects of interference and credibility on communication effectiveness. Psychological Reports, 1968, 23, 1123-1133.

Klatsky, R. L. Human memory: Structures and processes. San Francisco, Freeman, 1975.

Lindsay, P. H., & Norman, D. A. Human information processing, New York: Academic, 1977.

Mazzaro, D. W. Experimental psychology and information Chicago: Rand McNally, 1975.

McGuire, W. Inducing resistance to persuasion: Some contemporary approaches. In L. Berkowitz (ed.), Advances in Experimental Social Psychology, Vol. 1, New York: Academic Press, 1964, 191-229.

McGuire, W. Attitudes and opinions. Annual Review of Psychology, 1966, 17, 475-514.

Norman, D. A. Memory and attention. New York: Wiley, 1976.

Osterhouse, R. A., & Brock, T. C. Distraction increases yielding to propaganda by inhibiting counter arguing. Journal of Personality and Social Psychology, 1970, 15, 344-358.

Regan, D. T., & Cheng, J. B. Distraction and attitude change. Journal of Experimental Social Psychology, 1973, 9, 138-147.

Rokeach, M. Beliefs, attitudes, and values. Washington: Jossey-Bass, 1968.

Rosenblatt, P. C. Persuasion as a function of varying amounts of distraction. Psychometric Science, 1966, 5, 85-86.

Saltiel, 3., & Woelful, J. Inertia in cognitive processes: The role of accumulated information in attitude change. Human Communication Research, 1975, 1, 333-344.

Silverthorne, C. P. & Mazmanian, L. The effects of heckling and media of presentation on the impact of a persuasive communication. Journal of Social Psychology, 1975, 96, 229-236.

Vohs, 3. L., & Garrett, R. L. Resistance to persuasion: An integrative framework. Public Opinion Quarterly, 1968, 32, 445-452.

Ware, D.. P., & Tucker, R. K. Heckling as distraction: An experimental study of its effect on source credibility. Speech Monographs, 1974, 41, 185-188.

Wyer, R. S. Cognitive organization and change: An information processing approach New York. Wiley, 1974.

Zimbardo, P., Snyder, M., Thomas, 3., Gold, A., & Gewertz, S. Modifying the impact of persuasive communications with external distractions Journal of Personality and Social Psychology, 1970, 16, 669-680.

Appendix I: Distraction's Variables

  1. Relevance: the distraction may or may not be relevant to the content of the persuasive message. If the distraction is relevant, it can be expected to be integrated into the primary persuasive message.
  2. Evaluative bias: relevant distractions may reinforce or denigrate the content of the persuasive message. If the bias is positive, the distraction will reinforce arguments as being good. If the bias is negative, the distract ion will cue the audience to potential counter arguments.
  3. Distraction strength: among irrelevant distractions, increasing distraction strength will first act to reduce counter argument and increase persuasion, and then act to reduce both comprehension and persuasion. This relation ship is summarized by an equation forwarded by Wyer (1974) and can be represented by a bell shaped curve.
  4. Perceptual focus: focusing audience attention on the message decreases the effect of distraction. Focusing audience attention on the distraction increases the effect of distraction.
  5. Message complexity: more complex messages are more vulnerable to the effects of distraction
  6. Topic familiarity: high topic familiarity makes a subject less vulnerable to the effects of distraction. Low topic familiarity makes counter argument reduction very difficult. There are few, if any, to disturb. The result is a change in the shape of the curve representing the relationship between strength of distraction and level of influence.
  7. Channel overlap: secondary messages delivered to the same sensory node to which the primary message is delivered may be less distracting than would be the case if delivery was to a different sensory node.
  8. Message similarity: increasing similarity between the irrelevant distraction and the persuasive message may be more distracting than would be the case if they were less similar.

Appendix II: Vocabulary for an Information Processing Theory of Distraction in Attitude Change