ABSTRACT THE EFFECTS OF TRANSFORMATIONAL COMPLEXITY AND LANGUAGE INTENSITY ON RECEIVER COMPREHENSION AND ATTITUDE CHANGE by John R. Baseheart The present study investigated the effects of transformational complexity and language intensity on the level of comprehension and amount of attitude change demonstrated by message recipients. Ninety-five subjects were administered a pretest attitude questionnaire and a pretest comprehension examination. Subjects were then stratified on the basis of their pretest comprehension exam scores and assigned randomly to one of four experimental treatment conditions. In the posttest session, each subject first read one of four eXperimental messages, each of which differed in the combination of level of transformational complexity and language intensity, after which they completed attitude ratings of the message topic and the alleged message source, as well as ratings of various message characteristics and the perceived intensity level of the experimental message. Transformational complexity and language intensity were dichotomized into high and low levels. High transformationally complex messages were operationalized by the use of passive, self-embedded, and nominalized syntactic constructions, while low transformationally complex messages were constructed using active, non-embedded, and non-nominalized active verb-forms. High language intensity messages were operationalized by the insertion of 175 lexical items (adverbs, adjectives, and verbs) of known high intensity ratings into the messages at various apprOpriate locations. The same procedure was Mix John R. Baseheart employed, using lexical items of low intensity ratings, for the manipulation of the low language intensity messages. Evaluation of the effects of the two independent language variables on the dependent measure of comprehension was assessed by the use of thirty-eight multiple choice-type examination items, the number of questions answered correctly being the criterion measure employed for purposes of analysis. Receiver attitude change, the second dependent measure employed, was measured by use of subjects' pre-test to posttest attitude change ratings of the topic on five, evaluative dimension, semantic differential-type scales. Subjects also rated the perceived intensity level of the experimental messages, the perceived credibility level of the alleged message source, and the experimental messages on the four message characteristic dimensions of logic, quality, readability, and clarity. Two interaction hypotheses were tested in the investigation. These hypotheses stated that both comprehension and attitude change would be significantly inhibited by messages high in transformational complexity, particularly when messages were also high in language intensity; conversely, that comprehension and attitude change would be significantly facilitated by low transformationally complex messages, particularly when such messages were low in language intensity. The primary analyses failed to support either of these hypotheses. Neither transformational complexity nor language intensity in connected discourse significantly influenced receiver comprehension or attitude change. The two desired levels of message intensity were successfully manipulated as evidenced by the subjects' ratings of the high intensity messages as signi- ficantly more strong, certain, and emphatic than subjects' ratings of the low intensity messages. John R. Baseheart The two manipulated language variables produced no significant effect on the credibility ratings of the alleged message source in terms of his perceived justness, fairness, qualification, or information level. However, messages containing high intensity lexical items produced significantly higher ratings of the source in terms of his aggressiveness and emphatic qualities than did messages containing low intensity lexical itemso The ratings by the subjects of the experimental messages on the four message characteristics of logic, quality, readability, and clarity also reflected no significant differences among the groups processing the different experimental messages. Thus, all messages, regardless of level of complexity or language intensity, were perceived as being equally logical, good, readable, and clear° In an effort to ferret out possible relationships that may have been obscured by the partial failure in the manipulation of the two levels of message com- gflexity, supplemental analyses of the subjects' comprehension, attitude change, and credibility rating scores were performed, using only those individuals' scores who perceived the message complexity manipulations as intended. Some minimal support for the major hypotheses was derived from these analyses. The analysis of subjects' comprehension scores revealed that indi- viduals receiving a low intensity message attained significantly higher compre- hension scores than did individuals receiving a high intensity message. Also, the supplemental analysis of receiver attitude change scores yielded results for the transformational complexity variable which, although not significant at the prescribed level of significance, nevertheless approached significance (p.<.10) Thus, somewhat greater attitude change occurred under conditions of low message complexity than under conditions of high message complexity. Supplemental analyses of the source credibility ratings yielded significant message complexity effects for both the Safety and Qualification dimensions of John R. Baseheart credibility. The alleged source of the low complexity messages was perceived as significantly more just, fair, qualified, and informed than was the source of the high complexity messages. For the credibility dimension of Dynamism, the results indicated that the alleged source of the high intensity message was perceived as significantly more aggressive and emphatic than was the source of the low intensity message. Finally, a correlation analysis of comprehension and attitude change scores yielded a significant negative correlation between these variables for the combined g‘oups of subjects: as level of comprehension increased, the amount of attitude change demonstrated by subjects decreased. A number of research extensions, suggested by the findings of the investi- gation, were discussed. Accepted by the faculty of the Department of Communication, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for the Doctor of Philosophy degree. 4%? mm.» Director of Thesis 9 J Guidance Committee: . arm—if 77mm gig/v: ,mamm THE EFFECTS OF TRANSFORMATIONAL COMPLEXITY AND LANGUAGE INTENSITY ON RECEIVER COMPREHENSION AND ATTITUDE CHANGE by John R. Baseheart A THESIS Submitted to Michigan State University In partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1969 TABLE OF CONTENTS 525:. CHAPTER I INTRODUCTION . . . . o . . . . . . . . . . o . . . . o . o . 1 II METHOD . . . . . . . . . . . . . . . . o . . . . . . . . . 2O Pretests . . . . . . . . . o . . . . . o a . o . . . . . . 20 Independent Variables . o . . . . . . o . . o . . . . . . 2l Dependent Variables . . . . . o . . Message Intensity Manipulation Check a . o . . . . . . . o 26 Subjects . . . . . . . . o . o . o . . . o . . . . . . . . 27 Procedures . o o . . . . . . . . . . . . . . o . . . . . . 27 O O 0 O O O 0 0 ° 0 O 25 III SULTS O O 0 O 6 O 0 O O 0 O 0 O O O O 0 O O O 0 0 0 0 0 0 29 Pretest . o o . . . . o . . . . o . o . . o o . . . . o . 29 Tests of the Hypotheses . . . . . . . . . . o . . . o . . 30 Message Intensity Manipulation Analyses . . . . . o . o . 32 Source Credibility Ratings . . . . . . . . . . . . . . . . 3H Message Characteristics Analyses . . . . . . . . . o . o . 38 Correlation of Attitude Change and Comprehension Scores . Ml Supplemental Analyses . . . . . . . . . . . . . o . . o . ul Pretest comprehension scores . . . . . . . . . . . . . . H2 Tests of hypotheses . . . . o . . . . . . . . . . . . . #3 Source credibility . . . . . . . o o . . . . . . . o . H7 IV CONCLUSIONS, DISCUSSION, AND IMPLICATIONS . o . . o e . o o 52 Conclusions . . o . o . . . . . . . . . . o . . o . . o o 52 Discussion . . . . . . . . . . . . . o . . . . . . . o o 56 Source Credibility . . . . . . . . . . . o . o . . . . . . 6l Attitude Change versus Comprehension . . . o . . o o . o . 62 Implications for Further Research . . . . . . . o o . o o 63 RLIFERBNCES O 0 O 0 O O O 0 0 O O O O O O O O 0 0 O O O O O 0 O O O 67 iii 10. ll. l2. l3. 1”. 15. 16c LIST OF TABLES Message data for the four eXperimental message versions . o 0 Summary table of means, standard deviations, and analysis of variance of pretest attitude scores . . . . . . . . . o . Summary table of means, standard deviations, and analysis of variance of comprehension test items answered correctly . Summary table of means, standard deviations, and analysis of variance for receiver attitude change scores . o o o o . . Summary table of means, standard deviations, and analysis of variance of perceived message intensity ratings . . o . 0 Differences between pairs of means for the message intenSity ratings 0 e o o o a o o e o o o e o o o a o o e o Mean ratings and standard deviations for the source credibility dimensions of Safety, Qualification, and Dynamism Analysis of variance summary table of mean Safety dimension ratings 0 O O O O O 0 O O O 0 O O 0 O 0 0 O 0 O 0 O O O 0 0 Analysis of variance summary table of mean Qualification dimension ratings . o . . o . . . . . . . . . . . o o o . . . Analysis of variance summary table of mean Dynamism dimension ratings . . . . . o . . o . . . . . . o . . . . o . Differences between pairs of means for the Dynamism dimension of source credibility . . . o . o . . . . o . . o . Mean ratings and standard deviations for the message characteristic dimension of logic, quality, readability and Clarity O O O O 0 I C O O O O 0 O O O O O O 0 O 0 O O 0 0 Analysis of variance summary table of mean rating for the logical-illogical message characteristic . o . . o . . . . . Analysis of variance summary table of mean ratings for the Quality message characteristic . . . . . . o . . . . o o o . Analysis of variance summary table of mean ratings for the readability message characteristic . . . o . o o o . . . o . Analysis of variance summary table of mean ratings for the clear-vague message characteristic . . o . . . . . . . . . . iv 24 29 31 32 33 34 35 36 36 37 37 38 39 39 NO NO TABLE Page 170 Correlation between comprehension scores and attitude change scoms 0 O O O 0 C O O 0 O 0 O O O O O O O O O O O 0 O 6 ‘41 18. Means, standard deviations, and analysis of variance summary of pretest comprehension scores: Supplemental analysis . . . . H3 19. Means, standard deviations, and analysis of variance summary for treatment subgroup comprehension scores: Supplemental arlalYSiS O O 0 0 O O O O O O O O O 0 O O O O o O O O O 0 0 O 0 “1+ 20. Summary table of differences between pairs of means for subgroup comprehension scores . . . . . o . . . . . . o o . . 0 #5 21. Means, standard deviations, and analysis of variance summary for subgroup attitude change scores: Supplemental analysis . o 46 22. 'Mean ratings and standard deviations for the source credibility dimensions of Safety, Qualification, and Dynamism: Supplemental analysis 0 . . . . . . . . . . . . . . o . o . . . 47 23. Analysis of variance summary table for mean Safety dimension ratings 0 0 O 0 O O 0 O 0 O O O O O 0 I O O O O O O 0 0 O 0 0 1+8 2%. Analysis of differences between pairs of means for Safety Simention O O O 0 O O 0 O O O O O O O O O O O O O 0 0 O 0 O O O “8 25. Analysis of variance summary table for mean Qualification dimension ratings . . . . . . . . . . . . . . . . . . . . . . . #9 26. Analysis of differences between pairs of means for the Qualification dimenSion 9 0 O O O O O O O O O O O O O O O O 0 D 50 27. Analysis of variance summary table for mean Dynamicm dimension ratings . . . . . . . . . o a . . . . . o o . . . . o 50 28. Analysis of differences between pairs of means for the Dynamism dimenSj-on O O O O O O O O O O O O O 0 O O 0 0 O 0 O 0 51 29. Summary table of Flesch readability scores and comprehension scores for the primary and supplemental analyses 0 . o . . . . 58 APPENDIX A: APPENDIX B: APPENDIX C: APPENDIX D: APPENDIX E: LIST OF APPENDICES PRETEST ATTITUDE QUESTIONNAIRE . . PRETEST COMPREHENSION EXAMINATION EXPERIMENTAL MESSAGES . . . o . POSTTEST ATTITUDE QUESTIONNAIRE MESSAGE INTENSITY RATINGS MESSAGE CHARACTERISTIC RATINGS SOURCE CREDIBILITY RATINGS . . . POSTTEST COMPREHENSION EXAMINATION vi 0 0 103 106 v 1" CHAPTER I INTRODUCTION For the most part, persuasion researchers have concentrated on the effects of various personality and message variables on the amount of re- ceiver attitude formation and change toward various topics and message sources (for example, see Rosnow and Robinson, 1967, Chapters 1, 2, and 3). Minimal research has focused on the possible effects of various language variables on receiver attitude change. The research that has thus far been conducted has focused primarily on the influence of language intensity on recipient attitude change. The purpose of the present investigation was to investigate the possible effects of two language variables (transformational complexity of language and language intensity) on recipient comprehension level and degree of attitude change. Transformational Complexigi The effects of transformational complexity of language on language users has been attended to almost exclusively by linguists and psycholinguists, with the majority of the research concerned with the effects of syntactic structure on recall and comprehension of information. Undoubtedly the work of Chomsky (1957) in transformational generative grammar has stimulated the most pervasive interest in this area. As proposed by Chomsky, a transformational generative grammar is a finite grammar which Specifies a minimum set of rules for generating an infinitely \ - large number of grammatical sentences, while avoiding the formulation of any non-grammatical sentences; and which can assign a structural description to each sentence. This type of grammar consists of three major components: a syntactic component, a phonological component, and a semantic component. The syntactic component Specifies the deep structure of a sentence, for which the semantic interpretation is specified, and via transformations, the surface structure, for which a phonetic interpretation is determined by the phonetic component. It is the semantic component that assigns an interpretation to the deep structure underlying a sentence, whereas the phonological component assigns a phonetic iriterpretation to the surface structure. If the only transformational rules applied in generating a sentence are those designated as obligatory (e.g., trans formations determining the proper ordering of affixes and auxiliaries) a simple declarative, active sentence, with simple noun and verb phrases is generated (i.e., the kernel sentence). On the other hand, application of optional transformations to the strings underlying such sentences results in such derived syntactic forms as the passive, negative, queries, and so ono The greater the number of Optional transformations required to transform a kernel sentence into some other syntactic form, the greater the complexity of the particular trans- formed construction. For example, whereas the active to passive transformation requires the application of one rule of transformation, the active to negative- passive transformation requires the application of two, and thus is transfor- mationally more complex than the active to negative or active to passive trans- formations. The appearance of the transformational grammar approach in the late 1950's stimulated a great deal of research concerned primarily with the re‘ lationship of a number of syntactic structures, either singly or conjointly, to information processing as indexed by comprehension and recall of information. However, amount of attitude change produced as a function of variations in the transformational complexity of language structure within a persuasive message has been totally neglected by persuasion researchers. The present investigation was, then, an attempt to provide some answers to questions concerning the re- lationship of transformational complexity to recipient attitude change. Language Intensity_ A second area of concern in the investigation was that of language (message) intensity and its possible effects on comprehension and attitude cnegys To date minimal attention has been given to language intensity in persuasion research, and the reported research findings are too equivocal to allow for meaningful generalization. Whereas the above mentioned linguistic research concerned with transformational complexity has totally neglected the possible persuasive effects of various linguistic structures, language intensity resear h has totally neglected the possible effects of language intensity on the com« prehension of information. Perhaps the equivocal findings concerning attitude change and language intensity are a consequence of deficiencies in the compre- hension of the persuasive messages. If so, and if the assumption of Hovland, Janis, and Kelly (Insko, 1967) that comprehension is an important variable in the acquisition of new opinions is tenable, then such comprehension deficiencies may provide an eXplanation for the conflicting findings of investigations cone Icerned with these variables. Language intensity has been conceptualized and operationalized in at least two ways. First, Bowers (1963) states that language intensity can be Conceptualized as: the quality of language which indicates the degree to which the Speaker's attitude toward a concept deviates from neutrality. High intensity, thus, is characterized by emotionalism and extremity. This conceptualization was operationalized as follows: Bowers had subjects rate the perceived intensity level of either three or four lexical items that would be inserted into a particular position or location within a message (such anchors as "slightly intense," "quite intense," and "extremely intense" were employed in rating the lexical items). These items were rated within the context of the messages themselves by the subjects, which amounted essentially to a process of the subjects making comparative judgments of all the lexical items associated vvith any Specific location in the message. The items typically rated for in: tensity level were adverbs, adjectives, and verbs. The lexical items receiving the highest mean intensity ratings were inserted into messages, these messages being designated as high_intensity messages. The same procedure was followed for the construction of the low intensity messages (see Bowers, 1963, for a complete description of the methodology). A second method for manipulating language intensity in messages is one in which various lexical items of known intensity levels (usually adverbs, adjectives, and verbs) are inserted into persuasive messages. The intensity values for these lexical items have previously been established empirically, usually via various scaling methods or procedures, and thus have numerical intensity values assigned to them (see, for example, Howe , 1962; I966; Cliff, 1959; Dixon and Dixon, l96u; and Osgood, Saporta, and Nunnally, l956). The high and low intensity persuasive communications are then constructed by selecting items with either high or low known intensity values and inserting them into the messages at appropriate points. Relevant Research Transfvrmational Complexity In the present investigation, transformational complexity was defined by the use of variations within messages of three syntactic structures. The particular syntactic structures manipulated were active-passive voice, self— emheddedness versus non-embeddedness, and active verb forms versus nominalizations of these forms. Justification for consideration of various of these constructions as more complex than others will shortly become evident. Active-Passive. Of all the syntactical constructions investigated in the study of linguistic behavior, the active-passive voice construction hag pro- Chced the most consistent results in terms of the effects of these constructions on receiver'recall and comprehension. One of the earliest investigations employing this construction is reported by McKean, Slobin, and Miller (1962). These investigators assumed that he more cnmplicated the grammatical transformation, the longer it would take subjecrs to perfenn it. They gave subjects the task of pairing two sentences of differing syntactic types (e.g., a kernel and a passive er a passive and a negative, and so on) in a given length of time. The results indicated that passive constructions took longer to transform than kernel (active) sentences. In terms of correct recall of words in active and passive constructions, Epstein (1967) employed both structurally anomalous sentences and unstructured series of words in active and passive construetiou farms. He found that words in active cotstructions were recalled significantly better the? There contained in passive conr*ru;t§ Is. Coleman (1965) obtained similar reagltc using words, word sets, and complete sentences as response units. Also, as length of response units increased from words to word sets to sentences, the differences in recall between the transformation became more pronounced. When sentence recall is used as the dependent measure, the findings of a number of investigations demonstrate the generally greater facility of recalling active over passive constructions. Mehler (1963), in a prompted sentence recall 'task, found that kernel sentences were recalled significantly better than any other syntactic form, including the passive form. Savin and Perchonoch (1965), were concerned with a measure of the amount of immediate memory that is occupied by each of eleven different grammatical constructions, among which were kernel and passive constructions. Capacity for immediate memory for each of the grammatical types was operationalized as the total number of words recalled correctly 2ftgr_perfect recall of the designated sentence. In terms of the mean number of words correctly recalled after correct recall of the sentence types, significantly more words were recalled after a kernel sentence than after a passive sentence. Thus, it was posited that active sentences occupy less Space in immediate memory than passive ones, due to the lesser degree of transformational complexity of the active construction. Finally, Turner and Rommetveit (1968) manipulated the voice in which children recalled sentences by varying the child's focus of attention (on the actor or acted-upon object) at the time of sentence storage and retrieval. The finding of primary signifi- cance here is that children recalled active voice sentences significantly more frequently than passive voice ones. Comprehension as a dependent measure in the processing of active and passive constructions has generally taken the form of either multiple-choice tests or verification tasks. On a verification task, the subject is presented with a sentence of Specific grammatical type (e.g., passive-negative), followed by a stimulus (generally a picture) depicting events which confirm or falsify the sentence. The amount of time it takes subjects to decide upon the truth or falsity of the sentence with respect to the picture is the dependent measure assumed to reflect speed of comprehension or understanding. Coleman (196%) used two long prose passages which were Simplified by applying three transformations to them, one of which was the transformation of passive verb constructions to their active verb counterparts. Significantly more multiple-choice questions about the simplified version (the version con- taining the active verbs) were correctly answered than about the original version (the version containing passive verbs). Interpretation favoring the greater complexity of the passive construction must be tempered, however, due to the fact that two other grammatical transformations were also manipulated in these passages. Gough (1965) employed a verification task using active and passive con- structions. Active sentences were verified more rapidly than passives, again suggesting greater ease of comprehension of active sentences over passives. In a follow-up investigation, Gough (1966) conducted two experiments. The first one delayed the presentation of evidence confirming or disconfirming a sentence by three seconds after sentence presentation, and the second experiment con- trolled for differing lengths of active and passive sentences by deleting the agent phrase from the passive sentences. In both eXperiments, active statements were verified more rapidly than passive ones. Slobin (1966) employed a sentence verification task with kernels, passives, negatives, and passive-negatives, in which the pictures used as stimuli were either reversible or non-reversible (e.g., reversible, a boy hitting a girl; non-reversible, a dog biting a man). He found that when reversible sentences were used, active sentences were verified faster than passives, but when nonw reversible sentences were used, differences between active and passive constructimc essentially disappeared. Turner and Rommetveit (1967) also employed reversible eand non-reversible stimuli in investigating the ability of children to imitate, comprehend, and produce active and passive sentences. They found that active sentences were correctly reSponded to more frequently than passives, and none reversibles more frequently than reversible sentences. However, the effect of sentence voice was found to be stronger than the effect of reversibility. The order of comprehension difficulty from least to most difficult was: non~revers;k' active < reversible, active4< noncreversible, passive ( reversible, passive. Huttenlocher, Eisenberg, and Strauss (1968) report an investigation USILS time to perform a motor task as a dependent measure of the comprehensibility cf active and passive sentence constructions. Subjects were required to place cnn toy truck relative to a second truck (which was fixed in place), the instrucriani for placement being in either the active or passive voice. In all cases it teal significantly longer to place the movable truck when passive constructions were used, suggesting the greater difficulty in processing passive constructions. From the results of these investigations, it is apparent that active sentences are more easily processed, recalled, and comprehended by language UL645 than are passive constructions. Self—embeddedness. Far less evidence concerned with the effects of self embeddedness on information processingisawgfilableo A few investigations concernec with self-embeddedness and language processing are, however, available. Operationally, a construction is said to be self‘embedded when one lie guistic construction is inserted into the middle of another construction, with .aI ‘ the inserted construction being of the same grammatical form as the construction into which it has been inserted. For example, the sentence "The boy that the man saw stole the car" is an example of embedded sentence with one degree of self- embeddedness. In terms of the Speaker—hearer's ability to process these types of constructions, Miller (1962) suggests that self-embedded constructionsin language are more diffficult to process and to understand than either left- or right~ recursive constructions since the self-embedded constructions: by their very nature place heavier demands on the temporary storage capacity of any device that attempts to cope with it-- far heavier than do either left-recursive or right-recursive constructions. And Since our temporary memory is quite limited (see Miller, 1956), we can experience great difficulty following grammatical rules in this type of syntactic structure. Miller (1962) reports example of the difficulties subjects encounter when endeavoring to process and produce replications of multiple embedded sentences. Formalizing these early observations, Miller and Isard (1964) performed an investigation in which they employed sentences with zero, one, two, three, and four degrees of self-embeddedness, and random strings of words. Sentence recall was the dependent measure of concern, and was defined as the number of words recalled in the original sentence order. It was found that self‘embedding made sentences more difficult to learn; that on every trial, with the exception of one, the number of recall errors increased as the degree of self-embeddedness increased. In an investigation concerned with less highly embedded sentences such a. cane would eXpect to encounter more often in prose material, Coleman (1965) in~ vestigated self-embedded sentences containing only one degree of self-embeddedness. For example, the non—embedded sentence, "I gave the ball to the boy who lives here," had as its self-embedded counterpart the sentence, "I gave the boy who l.ives here the ball." (Coleman, 1965). The number of words correctly inserted 10 into the sentences using a cloze procedure technique was used as the dependent measure. It was found that significantly more words were correctly inserted into the non-self-embedded sentences than into the self-embedded sentences. Blumenthal (1966), using both errors and introspective data from subjects to understand how individuals process and comprehend multiple self-embedded sentences, employed sentences with three degrees of self4embeddedness in a comprehension task. Subjects rewrote or paraphrased sentences so as to make them more comprehensible or acceptable. The results indicated that in multiple self-embedded sentences, when subjects encounter successive noun phrases intro- duced by relative pronouns, these noun phrases are referred back to the original noun-subject rather than to each preceding noun. Thus, it was suggested that difficulty with the processing of multiple self-embedded sentence was with the assignment of grammatical relations to various sentence constituents. Further, the subjects actually perceived the multiple embedded sentences as ungrammatical approximations to sentences with only one embedding. Podor and Garrett (1967) conducted five related experiments concerned with various aspects of self-embeddedness and language processing. Their stimulus sentences were all double embedded constructions. The general eXperiment employed one list of sentences containing two relative pronouns per sentence used to introduce the two embedded noun phrases. A second list of sentences containing no relative pronouns before the embedded noun phrases and a third list of sentences containing adjectives to introduce the noun phrases (with the relative pronouns absent) were also used. The task of the subjects' was to paraphrase what they heard or saw. Performance was assessed on the basis of the mean number of grammatical relations (i.e., subject-object relations) correctly recovered per second cf reSponse delay. The results of primary interest here are 11 that the presence of the relative pronouns introducing the self-embedded noun phrases made sentences significantly easier to understand than when these pronouns were absent . These investigations reflect the consistent finding that self-embeddedness in language structure makes information processing significantly more difficult than language not containing self-embedded constructions. Nominalizations. A final class of syntactic transformations to be reviewed is that of the nominalization of active verb constructions: the grammatical area of English in which new nouns are created by the application of transforma- tional rules to active verb forms. For example, the sentence, "He explained the design," is the active verb construction of the nominalized form, "His explanation of the design." Although somewhat scanty, the relevant research nevertheless presents rather consistent results concerning the effect of nominalizations on recall and comprehension of information (see Lees, 1960 for a complete dis. cussion of nominalizations). Coleman and Blumenfeld (1963) reported an investigation in which nominalized sentences were compared by means of a cloze test to their grammatical trans- formations using active verbs. It was found that the mean number of content words correctly filled in for nominalized sentences was significantly lower than for active verb sentences. Coleman (1964) conducted a series of experiments concerned with the com- prehensibility of different grammatical transformations in a passage. In two experiments in which difficult prose material was simplified by transforming nominalizations, adjectivalizations, and passive verbs to their active verb forms, Coleman found that the active verb transforms were significantly easier to compre- hend than the original versions. In two other experiments concerned exclusively 12 with the effects on sentence recall of transformations of nominalizations to their active verb constructions, active verb transforms were recalled signifi- cantly better and took fewer trials to learn than nominalizations. Coleman (1965) further investigated the effects of nominalizations on recall by comparing ten different kinds of nominalized sentences to their de‘ transformed versions using active verbs, each kind of nominalization being represented by several sentences. He found, as before, that active verb trans- fo*mations were significantly easier to recall than their nominalizations. Epstein (1967) investigated the effect of active, passive, and nominalized sentences in recall. Using six anomolous and six unstructured series of words, he found that the mean number of words correctly recalled was significantly poorer for nominalized sentences than for either active or passive sentences. A final investigation is of particular relevance, since it incorporated a combination of four of these syntactic constructions and their transformations into prose passages and assessed their effects on recall and comprehension of prose material. Okatcha (1968) constructed high and low difficulty versions of garose messages and presented these versions to individuals under conditions of massed, distributed, and slow one-trial modes of presentation. Of particular interest here is the operationalization of the high and low difficulty messages. In the high difficulty messages, whenever possible, active sentences were changed to passives, non-embedded constructions were transformed to embedded versions, active verb sentence forms were transformed into various nominalized counter- parts, and noun clauses were included in the messages. Also, wherever possible, these transformed versions were combined into long complex sentences. The low difficulty versions thus contained active voice, non-embedded, simple sentence constructions. The results of the investigation indicated that significantly " l3 poorer recall and comprehension was produced by prose material containing more complex grammatical constructions (i.e., the high difficulty messages). Also, the slow one-trial presentation method facilitated recall and comprehension of the grammatically complex material. The results of the investigations reviewed above indicate the justification for the definition of levels of transformational complexity employed in the study. Language Intensity As mentioned above, the literature dealing with the effects of language intensity has dealt primarily with attitude formation and change and has produced results that do not allow for the formulation of any consistant and reliable conclusions. Comprehension and recall as dependent measures of language intensity research have not been attended to by persuasion researchers. Bowers (1963; 196M) and his associate (Bowers and Osborn, 1966) have shown the greatest interest to date in the effects of language intensity on receiver attitude change. Operationalizing the concept of language intensity as described above, Bowers (1963) investigated the relationship between language intensity, social introversion, and attitude change. He hypothesized that high intensity messages would produce greater attitude change toward both the concept and the speakers than would low intensity messages. The data failed to support either of these hypotheses. Bowers did obtain a significant language intensity by direction of attempted influence interaction, indicating that intensity had a differential effect on Speakers arguing in favor of, or opposed to, an issue. It was found that lgw_intensity anti- messages were significantly more effective in changing attitudes than were the corresponding high intensity messages -- a reversal of the expected relationship between these variables. Bowers' explanation for this - n. I . 1.- . 1H phenomenon was that perhaps the extremity of word choice produced a "boomerang" effect, and that the various highly loaded constructions may have produced a credibility gap toward the Speaker, thus adversely affecting receiver's attitudes. Bowers (1964) endeavored to isolate some correlates of language intensity so that ultimately one might be able to predict the intensity level of a lexical iten from various other attributes of that item. Using correlational analyses, Bowers found: (1) a significant but low correlation (3b = .10) between intensity and item length in syllables; (2) a moderate correlation (Eb = .59) between how obscure a term was judged to be and intensity level; (3) a high correlation (Eb = .89) between the presence of qualifying terms and intensity, and (u) a high correlation (Eb = .83) between language intensity and metaphorical quality of two types of metaphors (i.e., a death and a sex metaphor). The findings of the apparent strong relationship between language intensity and metaphorical quality led Bowers and Osborn (1966) to conduct an investigation of the effects of concluding metaphors on receiver attitudes toward the messages and the credibility of the sources. Two metaphors were employed: a sex metaphor and an extended death metaphor. These messages were attributed to either an economics or a science professor. The results indicated that, for both Speeches, more attitude change was produced in the intended direction by intense metaphorical conclusions than by literal conclusions. The hypothesized relationship between intense metaphorical quality and the credibility of the source was confirmed, although the relationship was more complex than was ex- pected for the various credibility dimensions used. A final investigation which employed Bowers' operationalization of language intensity was reported by Carmichael and Cronkhite (1965). These authors took a suggestion from Bowers' (1963) investigation that the unexpected findings that 15 intense language produced less attitude change than non-intense language may have been produced by the inadvertent frustration of experimental subjects by the experimental manipulations. Thus, these authors hypothesized that frustrated subjects would show a greater tendency to agree with Speakers using low intensity language than with Speakers using high intensity language, whereas this pattern would not occur for the non—frustrated subjects. Frustration was conceptualized as the blocking of behavior directed toward the attainment of some desired goal. That is, frustration was assumed to be produced in the presence of two criteria: (1) the existence of a desired goal reSponse for the subject, and (2) the blocks ing of this response (for a discussion of this position, see Dollard, J., g: 31, 1939). The major hypothesis was confirmed: frustrated subjects who heard messages containing low intensity language were more favorable toward the topic of the Speech than were frustrated subjects who heard high intensity Speeches. This pattern, as predicted, did not occur for the non-frustrated subjects. The I~esults of this investigation are relevant to the discussion of theoretic con: siderations of language intensity and transformational complexity presented later in this discussion. A final investigation of relevance is reported by Kochevar (1967)° This investigation attempted to measure the effects of varying the levels of message intensity on receiver's attitude toward the message, the topic, and the source of the message. Message intensity was operationalized by the second method mentioned earlier; that is, Kochevar constructed his messages using previously rated high and low intensity adverbs, adjectives, and verbs. The results clearly indicated that the attempt to manipulate varying levels of message intensity was 16 successful; however, none of the main hypotheses was confirmed. When compared to low intensity messages, high intensity messages did not produce greater attitude change toward the topic, source, or message characteristics. In general, then, high and low intensity messages did not differentially affect receiver attitudes. The conflicting findings of these message intensity investigations point out the present impossibility of formulating any low-level generalizations con» ceerning the isolated effects of varying levels of language intensity on receiver attitude change. Perhaps, as Bowers has suggested: "... language intensity ... must be regarded as a complex variable which is subject to interactions with many other variables." (Bowers, p. 352, 1963). To date, only Carmichael and Cronkhite have attempted to explicate the possible interactive effects of language intensity with other variables. Thus, the intent of the present investigation was to investigate the effects of language intensity somewhat further, this time as it may interact significantly with message complexity to influence the level of recipient information comprehension and amount of attitude change. hypotheses Consistent with the above remarks of Bowers, and particularly in light of the findings reported by Carmichael and Cronkhite concerning the interactive effects of frustration and language intensity on receivers' agreement with an advocated position, an interactive relationship between language intensity and transformational complexity of language was hypothesized in this study: Hl: When message complexity is high, subjects' informa- tion comprehension will be low, especially when message intensity is also high; when message complexity is low, however, subjects' information comprehension will be high, particularly when message intensity is lowo 17 This hypothesis Specified an interaction between level of message intensity and degree of transformational complexity. The rationale for this hypothesis was as follows: Given the transformational complexity findings and recalling the Carmichael and Cronkhite investigation concerned with the effects of frustra- tion on receiver's attitudes, it can be suggested that messages of high trans- formational complexity will produce some increment in frustration (in the Carmichael and Cronkhite sense) experienced by subjects. For since highly complex messages are significantly more difficult to process and to comprehend, the desired goal of achieving a maximum comprehension score is to some extent blocked. The blocking of this goal should in turn increase the individual's frustration level. If this increased frustration does occur, then individuals receiving a highly complex message should comprehend more information under conditions of low message intensity than high message intensity. This conclusion is supported by the Carmichael and Cronkhite investigation in which high frustration produced a decrement in performance when coupled with high message intensity and an increment in performance when coupled with low message intensity. However, for low complexity messages, greater comprehension should occur under conditions of low rather than high message intensity. This relationship is suggested by the findings that in certain instances, low intense messages are more effective in modifying behavior than high intense messages. Further, low complexity messages should significantly facilitate comprehension over that produced by high com» plexity messages, regardless of level of message intensity. Following from Hypothesis 1, if as suggested earlier, the relationship between comprehension and attitude change is a positive one (an assumption which, although somewhat debatable, draws some support from investigations by Greenberg, 196a; Insko, 196u; and Watts and McGuire, 196u), it would be expected that amount 18 of attitude change produced as a function of transformational complexity and language intensity would follow the same pattern as delineated for comprehension. Thus, Hypothesis 2 stated: H2: When message complexity is high, subjects will change their attitudes toward the message topic less, eSpecially when message intensity is also high; when message complexity is low, however, subjects will change their attitudes toward the message topic more, particularly when message intensity is low. The rank ordering of the complexity by message intensity conditions (from least to most comprehension and attitude change) would thus be as follows: high complexity, high intensity‘high complexity, low intensitye some proof that in Massachusetts democracy seems to work and that their best t alent for leadership is sometimes drawn from the rank and file of peOple. Mr. Kennedy won. Further, party leaders' attempts to offer at least a somewhat "balanced ticket" are sometimes defeated by ethnic minorities grouped in parties, thus jprobably dooming to possible defeat their whole ticket in the general election. Also, rather reSponsible candidates for public office are sometimes defeated by palpable demagogues. That the use of a direct primary probably may not be cbsirable seems to be suggested by all of these results. We perhaps ought not, if we possibly want political parties, cut them off from the process of choosing candidates for public office. They probably should not be deprived of motives 1:0 organize. They should not sometimes be set by us prematurely at the mercy of hordes of peOple whose primary stage knowledge may be somewhat poor. That this may not be an elitist doctrine seems Somewhat apparent. The fact probably should be faced by reSponsible political analysts and advocates that the safe cognitive anchorage around which political preferences are organized may conceivably be provided by party identification for people. Set partially adrift from this anchorage, as they seemingly are when faced with an intra-party primary election, most voters have perhaps little or nothing to guide their choices. Many voters are probably given the only clues to choice by chance knowledge about a famous name or stray feeling of ethnic kinship under these conditions. Thus, probably one should, considering the somewhat low level of pOpular 92 interest and participation in political activities which seemingly prevails today, thoughtfully question the throwing of the future of the party system even partially and perilously into hands of primary electorates. It seems somewhat clear that the establishment of a national popular-vote primary as a means for selecting Presidential candidates at the present time may possibly be infeasible and undesirable, and therefore, should probably not be adopted. 93 LOW COMPLEXITY, HIGH INTENSITY MESSAGE A National Popular-Vote Primary? The past Presidential election has definitely given rise to some objections about the way in which candidates are chosen. Persons continually prOposing the use of a popular-vote primary on a national basis as a means for choosing Presidential candidates have often repeatedly voiced cries for change in the selection of candidates. This method looks desirable on the surface of things. However, there are definitely disadvantages with this method which render it positively untenable as a viable means for choosing Presidential candidates. First, it is extremely likely that as many as ten candidates will definitely get enough names on nominating petitions to get on the ballot. Nor is to too surprising if they Split the vote equally. This definitely denies any one person an absolute majority vote. The victor obviously then has to be chosen in a special run-off primary. Using this method, the United States must undeniably always restrict its prospective Presidential candidates to wealthy persons only. Anyone without great financial means could certainly never raise the millions required for the nominating petition, the first primary, the run-off primary, and the national election. No one who is not very superbly conditioned will survive the extremely demanding pace of all these campaigns. National primaries will definitely also produce an unquestionable weakening of the party system. It is very usual for a party to remain continually in office for a very long period of time. State experience with primaries is definitely a proven guide here. This results repeatedly in a decided movement of the very interested voters into the primary of the winning party where their individual votes unquestionably always count more. Doubtlessly as voters 99 continually desert the losing party, it is largely only the adamant die-hards who are left. They will obviously nominate candidates who definitely please them. These candidates, however, will undoubtedly not win the election Since they are definitely unappealing to the large masses in the country. Ultimately, the losing party will undoubtedly atrophy. This unquestionably weakens the two- party system decidedly and the prospects of competition among parties. The winning party definitely will soon Show Signs of decided internal weakness as a result of the complete lack of opposition required to keep it unified. A national primary will definitely produce the appearance of extremist candidates and demagogues. With absolutely no allegiance to any fixed party organization, they have undeniably little to lose by continually stirring up extreme mass hatreds or making positively wild promises. A Huey Long will un- questionably find a very fertile field in a national primary. This is obviously a chance sufficient enough to instantly raise the anger of American politics to unusually explosive levels even though he did not win. The convention system positively rules out these extremists. It definitely places the reSponsibility totally in the hands of decided party leaders who very definitely have a lasting stake in the keeping of the good name and integrity of their organization. An immediate view into this problem is had by looking at the situation in any of the Southern states. Here voters definitely vote only in the Democratic primary. Winning in that primary is unquestionably tantamount to elections. The result of this is an extremely wild, factional politics in which there are very few or no fixed party leaders. This completely blurrs the unquestionable distinction between the "ins" and the "outs". It is extremely hard with this to hold any- one totally responsible. Demagogues repeatedly arise who make decided use of this 95 situation by strident appeals. In some primary systems an extremist unquestionably always takes the place of the party. Thus, he always gives a type of minimal structure to the State politics. This must definitely give total pause to the advocates of a national primary. Thus, any use of the direct primaries definitely weakens the party system. Only the very richest candidates are able positively to enter a large number of them. They definitely encourage the prospective candidates to continually bypass all regular party organizations in favor of campaigns stressing completely personal publicity. Also, they undeniably throw the nominations totally into the hands of those whose stake in the workings of the political process is definitely not great enough to ensure that the final nominee is unquestionably qualified for the Presidency by qualifications of experience, qualities of mind, or political alliances with others professionally engaged in political actions. Data from the State level primaries definitely throws decided light on the prdblems involved in a national primary system. The use of primaries at the state level definitely produces abnormal experiences. For example, completely naive voters in primaries nominate totally unqualified candidates whose names resemble 11oted politicians. The nomination of John Kennedy by Massachusetts Democrats in 1959 as their candidate for state treasurer certainly shows one example of this. A minor clerical worker of a safety razor firm, Kennedy happened to have the same name as the junior United States Senator. The name of Kennedy is unquestionably strong medicine in the politics of Massachusetts. Kennedy won the primary nomination in a contest with the person the Democratic pre-primary convention endorsed. The Democratic State Committee swallowed its surprise. In its campaign literature it described Mr. Kennedy's unusual political success as "very American." It was said that Mr. Kennedy had obviously caught the sprit and interest of his 96 fellow citizens. His election was without doubt definite proof that in Massa- chusetts democracy certainly works and that Americans definitely always draw their best talent for leadership from the rank and file of the peOple. Mr. Kennedy won. Further, the ethnic minorities grouped in one party perpetually defeat the attempts by party leaders to offer totally "balanced tickets." Thus, the whole ticket in the general election is unquestionably doomed to toal defeat. Also, palpable demagogues repeatedly defeat very reSponsible candidates for public office. All of these results clearly Show that the use of a direct primary is definitely and unquestionably not desirable. Since we certainly want political parties, we positively must not cut them off from the process of choosing can- didates for public office. We definitely must not deprive them of motives to organize. We must certainly not repeatedly set them prematurely at the mercy of hordes of people whose knowledge level at the primary stage is extremely poor. It is undeniably apparent that this is not an elitist doctrine. Responsible political analysts and advocates must certainly face the fact that party identification for most people definitely provides the safe cognitive anchorage around which people certainly always organize political preferences. Most voters have very little or nothing to guide their choices when set completely adrift from this anchorage, as they definitely are when faced with a system of intra—party primary elections. Chance knowledge about a famous name or stray feelings of ethnic kinship under these conditions undeniably gives many voters the only clues to choice. Thus, definitely one must thoughtfully question the throwing of the future of the party system completely and perilously into the hands of primary electorates considering the extremely low level of popular interest and participation in 97 political activities which obviously prevail today. It is extremely clear that the establishment of a national popular-vote primary as a means for selecting Presidential candidates at this time is definitely and undeniable infeasible and undesirable, and therefore, must definitely not be adopted. 98 LOW COMPLEXITY, LOW INTENSITY MESSAGE A National Popular-Vote Primary? The past Presidential election has supposedly given rise to some objections about the way in which candidates are chosen. Persons proposing the use of a popular-vote primary on a national basis as a means for choosing Presidential candidates have occasionally voiced some cries for change in the selection of candidates. This method may seem desirable on the surface. However, there appears to be perhaps some disadvantages with it which render it possibly un- tenable as a means for choosing Presidential candidates. First, it seems somewhat likely that as many as ten candidates might possibly get enough names on nominating petitions to get on the ballot. Nor would it be surprising if they Split the vote equally. This could deny any one person a possible majority vote. The victor would then probably have to be chosen in a Special run-off primary. Using this method, the United States might conceivably eventually have to restrict its Presidential candidates to wealthy persons. Any- one without great financial means could probably seldom raise the millions re- quired for the nominating petition, the first primary, the run-off primary, and the national election. No one who was not rather superbly conditioned could survive the demanding pace of all these campaigns. National primaries might also produce perhaps a conceivable weakening of the party system. It seems somewhat usual for a party to ultimately remain in office for a fairly long period of time. State experience with primaries can perhaps be a guide here. This could result sometimes in a possible movement of partially interested voters into the primary of the winning party where their votes perhaps would sometimes count more. As voters eventually desert the losing uh VI' '0 l we.“ ‘uin'za. ,. 99 party, it would be largely a few die-herds who would be left. They would probably nominate candidates who seemingly pleased them. These candidates, however, could probably not win the election since they would probably be unappealing to the masses in the country. Ultimately, the losing party could perhaps atrophy. This possibly could weaken the two-party system and the prospects of competition among parties. The winning party would possibly soon Show signs of some internal weakness as a result of the partial lack of opposition required to keep it unified. A national primary might possibly produce the appearance of extremist can- didates and denagogues. With seemingly no allegiance to any fixed party organi- zation, they would have little toleee by occasionally Stirring up some mass hatreds or seemingly making wild promises. A Huey Long might perhaps find a somewhat fertile field in a national primary. This might conceivably be a chance sufficient to eventually raise the anger of American politics to somewhat ex- jfiosive levels even if he did not win. The convention system supposedly rules out these extremists. It seems to place responsibility partially atleast in the hands of several party leaders who seem to have a lasting stake in keeping the good name and integrity of the organization. Some partial view into this problem may be had by looking at the Situation in some Southern states. Here most voters prdbably vote only in the Democratic primary. Winning in that primary may perhaps be tantamount to election. The result may be a slightly wild, factional politics in which there are a few or no fixed party leaders. This may blurr somewhat the (zonceivable distinctions between the "ins" and the "cuts". It may be somewhat hard to hold anyone even partially responsible. Demagogues occasionally arise who make some use of this situation by strident appeals. In some primary systems 100 an extremist can perhaps take the place of the party. He can sometimes give a kind of minimal structure to the state politics. This Should perhaps give pause to the advocates of a national primary system. Thus, even some use of direct primaries would probably weaken the party system. Probably only the somewhat richest candidates could possibly enter a large number of these primaries. They would perhaps encourage prospective candidates to sometimes bypass various regular party organizations in favor of campaigns stressing somewhat personal publicity. Also, they possibly would throw nominations partially into the hands of those whose stake in the workings of the political process may not perhaps be great enough to ensure that the final nominee seems qualified for the Presidency by qualifications of experience, qualities of mind, or political alliances with others professionally engaged in political actions. Data from state level primaries may perhaps throw light on the prdblems in- volved in a national popular vote primary. The use of primaries at the state level seemingly produces abnormal experiences. For example, somewhat naive voters in the primaries nominate unqualified candidates whose names resemble noted politicians. The nomination of John Kennedy by Massachusetts Democrats in 1959 as their candidate for state treasurer seemingly shows one example of this. A minor clerical worker of a safety razor firm, Kennedy happened to have the same name as the junior United States Senator. The name of Kennedy seems to be somewhat strong medicine in Massachusetts politics. Kennedy won the primary nomination in a contest with the person the Democratic pro-primary convention endorsed. The Democratic State Committee swallowed its surprise. In its campaign literature it described Mr. Kennedy's ultimate political success as "very American." It was said that Mr. Kennedy had probably caught the Spirit and interest of his fellow citizens. 101 His election would probably be some proof that in Massachusetts democracy seems to work and that Americans sometimes draw their best talent for leadership from the rank and file of the people. Mr. Kennedy won. Further, ethnic minorities grouped in one party sometimes defeat attempts 13y party leaders to offer at least somewhat "balanced tickets." Thus, the whole party ticket in the general election seems doomed to possible defeat. Also, palpable demagogues sometimes defeat rather responsible candidates for public c>ffice. All of these results seem to suggest that the use of a direct primary probably may not be desirable. If we possibly want political parties, we perhaps ought not cut them off from the process of choosing candidates for public office. We probably should not deprive them of motives to organize. We probably should not sometimes set them prematurely at the mercy of hordes of people whose knowledge at the primary stage may be somewhat poor. It seems somewhat apparent that this may not be an elitist doctrine. Responsible political analysts and advocates probably should face the fact that party identification for most peOple conceivably provides the safe cognitive anchorage around which people sometimes organize political preferences. Most voters have perhaps little or nothing to guide their choices when set at least partially adrift from this anchorage, as they seemingly are when faced with an intra-party primary election. Chance knowledge about a famous name or stray feelings cf ethnic kinship under these conditions probably gives many voters the only clues to choice. Thus, probably one should thoughtfully question throwing the future of the party system even partially and perilously into the hands of primary electorates considering the somewhat low level of popular interest and participation in 102 political activities which seem to prevail today. It seems somewhat clear that the establishment of a national popular-vote primary system as a means for selecting Presidential candidates at this time may possibly be infeasible and undesirable, and therefore, should probably not be adopted. APPENDIX D POSTTEST ATTITUDE QUESTIONNAIRE MESSAGE INTENSITY RATINGS MESSAGE CHARACTERISTIC RATINGS SOURCE CREDIBILITY RATINGS 109 POSTTEST ATTITUDE QUESTIONNAIRE Now that you have read the article, we would like to know your attitude, as of right now, toward the topic discussed in the editorial. Please make one, and only one, mark on each scale. Please do not omit any scales. The establishment of a national popular-vote primary as a method of Ehoosing Presidential candidates. 'f El:1... Bad : : : : : : Good Very Quite Slightly Neutral Slightly Quite Very V fly" vmfinnrnfiwwwrw ll Honest : : : : : : Dishonest Very Quite Slightly Neutral Slightly Quite Very Wise : : . . : : Foolish Very Quite Slightly Neutral Slightly Quite Very Unfair : : z : : : Fair Very Quite Slightly Neutral Slightly Quite Very Favorable : 2 ' ° ' : Unfavorable Very Quite Slightly Neutral Slightly Quite Very { neeeeeeeenne Now we are interested in what you thought of the article. Please make a check mark according to how you feel about the editorial. I How strong or 333k was the general tone of the writer's statements? Strong : : : : : : Weak Very Quite Slightly Neutral Slightly Quite Very How uncertain or certain was the writer about what he was saying? Uncertain : : : : : : Certain Very Quite Slightly Neutral Slightly Quite Very How hesitant or emphatic was the language that the writer used? Hesitant : : : : : : Emphatic Very Quite Slightly Neutral Slightly Quite Very How would you rate the logic of the arguments presented? Logical : : : : : : Illogical Very Quite Slightly Neutral Slightly Quite Very 105 How would you rate the readaaility_of the editorial? Un- readable : z : : : : Readable Very Quite Slightly Neutral Slightly Quite Very How would you rate the clearness of writing of this editorial? Clear : : Vague Vm NC te Vlightly Neutral Slightly Quite Very How would you rate the qualiEy of the information in this article? Very Quite Slightly Neutral Slightly Quite Very nee:.*mtec Finally, we would like to know what you think of the writer of this editorial. Make only one mark on each scale, and do not skip any scales. How unjust or jg§£_do you think the writer is? Unjust : : : : : Just Very Quite Slightly Neutral Slightly Quite Very How ungualified or qualified do you think the writer is? Un- qualified : : : : : : Qualified How uninformed or informed do you think the writer is? Un- informed : : : : : : Informed Very Quite Slightly Neutral Slightly Quite Very How aggressive or meak do you think the writer is? Aggressive : : : : : : Meak Very Quite Slightly Neutral Slightly Quite Very How hesitant or emphatic do you think the writer is? Hesitant : : : : : : Emphatic Very Quite Slightly Neutral Slightly Quite Very **39 {5453.3 4" 3“ Have you made one check mark on every scale? When you are finished, please raise your hand. We will begin the second part very shortly. APPENDIX E POSTTEST COMPREHENSION EXAMINATION 107 INSTRUCTIONS On the following pages are a number of questions concerned with the content of the editorial you have just finished reading. Please answer ESEh_of these questions only on the IBM Scoring Sheet provided, using the pencil we have provided for you. Do not skip any questions, please. You will have a maximum of 30 minutes from the time you are instructed to begin to complete this part of the study. The only information you need put on the Scoring Sheet is your name, student number, and instructor's name. When you have completed this part of the study, please raise your hand and someone will collect your materials. We thank you very much for your help and cooperation on this project. DO NOT TURN THE PAGE UNTIL YOU ARE INSTRUCTED TO DO SO. 108 INSTRUCTIONS: Answer all of the following questions on the IBM Scoring Sheet provided. Please do not skip any questions. 1. One consequence of a national popular-vote primary would probably be 1) 2) 3) 9) As the term "atrophy" was used in the selection, it probably meant l) 2) 3) 9) According to the article, state level experiences with primaries suggests strengthening of the internal structure of the winning party. the possibility of the weakening of the two-party system. an increase in the competition among the various parties. all of the above are likely consequences. L I g ts-l. cohesiveness demagoguery degeneration the selection gives no clues to its meaning. “‘1' monk-1151.15 I '3‘." Wi. '4'- 'n u-‘Inu ‘, )- na 1: :— fi.’ that a national primary would quite possibly 1) 2) 3) 9) increase the participation of individuals in the general election. place the nominations into the hands of only a small number of individuals who are best qualified to judge the qualifications of the nominee for the Presidency. produce candidates who are more acceptable and desirable to the general voting electorate than the present convention system does. produce movement of the interested voters into the primary of the winning party. Which of the following probable consequences of a national popular-vote primary was not suggested in the article? 1) 2) 3) 9) perhaps as many as a dozen or so candidates could secure enough signatures on petitions to get on the ballot. party reSponsibility would be placed squarely in the hands of the party leaders. there could be an equal division of votes among a number of candidates. a degeneration of intra-party opposition within the winning party. In the selection, the term "demagogue" probably refers to 1) 2) 3) 9) individuals who resort to force and coersion upon occasion to achieve various political ambitions. individuals who restrict their allegiance to a single permanent party organization without regard to candidates or issues. individuals who use the passions and prejudices of people for their own political advancement. individuals who employ strong party identification as a means to sway voters to the party regardless of the consequences involved. 6. 10. 11. 109 A convention system for selecting Presidential candidates rules out the possible influence of extremists because 1) there are few, if any, permanent party leaders established. 2) the distinctions between the "ins" and the "cuts" becomes blurred. 3) the responsibility for the party is placed in the hands of party leaders. 9) none of the above are applicable. One might infer from the selection that strong identification with a particular political party by voters 1) serves as a cognitive anchor for the organization of political preferences. 2) tends to perpetuate straight ticket voting in direct primaries. 3) increases the voting strength of the ethnic minorities in a direct primary. 9) could result in the election of a completely unqualified candidate. According to the selection, various state level primaries have in the past produced all of the following anomalous eXperiences except 1) the emergence of a demagogue as Spokesman for the majority party. 2) totally unqualified candidates who have been nominated by the voters. 3) ethnic minorities have produced defeat for the entire party ticket in the general election. 9) extremist candidates have defeated responsible candidates for office. According to the article, Huey Long was an example of l) a type of candidate strong enough to withstand the rigors of a popular primary. 2) an explosive man in American politics. 3) the pro-convention favorite of the Massachusetts Democratic party for state treasurer. 9) a politician with little allegiance to any permanent party organization. The article points out that one means to guard against such occurrances as ethnic ties determining selection of a political candidate is 1) education of the electorate concerning the candidates. 2) strong party identification. 3) a campaign stressing extensive personal publicity by candidates. 9) a direct popular-vote primary. According to the article, a party generally remains in power for relatively long periods of time. Eventually interested voters move into the primary of this winning party. This leaves only the die-hards in the party not in power. From the above stated relationships, which of the following conclusions appears most warranted? l) the losing party would begin to Show signs of internal weakness. 2) a strong nucleus of minority party members could select a strong candidate to run for office. 3) the losing party would perhaps demonstrate some degree of degeneration. 9) none of the above conclusions appear warranted. 110 12. Without exception, a national popular-vote primary process would require 13. 19. 15. 16. all of the following except 1) a primary. 2) a national election. 3) a run-off primary. 9) nominating petitions. The winning party in a national pOpular primary would, according to the article. 1) ultimately Show signs of internal weakness. 2) ensure a stronger political system than would occur under a convention system. 3) easily acquire, generally, a disproportionately large majority of the popular vote . 9) all of the above would quite possibly occur. The example from the selection concerning a Mr. Kennedy winning the Democratic primary election in Massachusetts demonstrates one problem of direct primaries of l) educating the voters with respect to the qualifications and abilities of the candidates running for office. 2) how easily the general voting public can be influenced toward a particular party candidate by a clever engineering of a campaign strategy. 3) questionably qualified persons being nominated by innocent voters in a primary. 9) the influence of the most powerful political party in a state on the outcome of a primary election. The most deleterious effect on the losing political party under a system of a national popular-vote primary would be to produce 1) atrophy. 2) increased internal unity within the party. 3) stagnation. 9) an internal weakness within the party structure. In the direct primaries, one of the greatest dangers to the "balanced ticket" is l) palpable demagogues and extremists. 2) wealthy and influential candidates. 3) a totally unqualified candidate. 9) ethnic minority groups. 111 17. According to the selection, qualifications for the Presidency include all of the following except 1) 2) 3) 9) qualities of mind. political alliances. experience. physical stamina. 18. By the "ins" and "outs" referred to in the selection, the author probably meant 1) those individuals who are, and are not, members of a particular political party. 2) the particular parties presently in power and out of power. 3) the extremist candidates versus the non-extremist candidates. 9) the party leaders of the majority party versus the party leaders of the 19. One is 1) 2) 3) 9) minority party. disadvantage of a national primary system pointed out in the article that possible candidacy would be restricted to only the more affluent upper-middle and upper economic classes of individuals. that regular party organizations and procedures would be bypassed in favor of personal publicity. that party responsibility would be placed in the hands of a number of party leaders. none of the above are significant disadvantages. 20. According to the selection, extremist candidates under a national primary system 1) 2) 3) 9) could possibly produce some degree of cognitive anchorage, and hence party identification, for a significant number of voters. might actually take the place of the party by giving structure to the political system. would quite possibly upset the "balanced ticket" of a party, resulting quite likely in defeat of the general party ticket. all of the above possibilities could be produced by an extremist candidate. *eeeeeeeaeeaeeeeeeee Mark the fellowing items according to the following key: KEY: l. A statement which agrees with the arguments and information 21. One presented. 2. A statement which disagrees with the arguments and information presented. 3. A statement which has no relevance to, or cannot be logically drawn from, the arguments and information presented. major advantage of a direct popular primary is that it is very unlikely that an extremist candidate could dupe enough of the voters to defeat the more responsible party candidates. 22. 23. 29. 25. 26. 27. 28. 29. 30° 31. 32. 33. 39. 112 One undesirable consequence of a national primary would be the weakening of both the intra-party as well as inter-party competition within and among the various political parties. In the convention system for nominating candidates, it is often the case that unattainable promises to the voters concerning policies and issues are made by candidates. One possible disadvantage of a national primary is that the winning party would quite probably Show Signs of internal weakness and deterioration over a period of time. A direct state primary system has the advantage of making it difficult for extremists to assume control of the party and hence give structure to the state '3 politics . In the case of an intra-party primary election, most voters have little or nothing to guide their choices since they have little party identification. Huey Long is an example of an extremist who, through a state level popular vote primary, gained and controlled a Southern State's politics for many years. National popular-vote primaries would be to the advantage of the more educated electorate since they would know more about the various candidates. The present convention system used by some states for selecting a Presidential candidate is particularly susceptible to the influence of extremists and demagogues. It would be extremely dangerous to put the selection of Presidential candidates into the hands of an electorate who have been shown conclusively to be highly uninformed about party policies, as well as candidates, of various political organizations. An advantage of the present state level primary system is that it allows for qualified persons to be nominated. Mr. Kennedy, mentioned in the article, won the primary nomination for Democratic state senator against the pre-primary convention candidate endorsed by the Democratic party of that state. One advantage of the present convention and state primary system is that both upper-middle and upper class individuals are able to afford the financial expense involved in waging a political campaign. The recent Presidential primaries clearly point out the need in this country for a revision of the present system of selecting Presidential candidates. 35. 36. 37. 38. 113 In various Southern states, victory in a Democratic primary is the same as election, this process producing a stable and orderly party system with relatively permanent leadership. In the event that a clear majority vote is not received by any candidate in a national primary election, individual state run-off primaries would be necessary to limit the number of candidates prior to a national run-off primary election. Extremist candidates, having strong allegiances to their party organizations, may present such a unified party front that nomination via a direct primary on a national basis could be a very likely occurrance. One of the greatest assets to any party in a national popular primary is the participation of various ethnic minority groups who tend to engage in "block" voting. STATE UNIV. LI nrcwrcnw lll‘lWl‘lNllm 3129 BRQRIES ll 1“ HWIHI WIHHHII (hi3 3006664fl