Aspects of Literary Response: A New Questionnaire
David S. Miall and Don Kuiken
Departments of English and Psychology
University of AlbertaResearch in the Teaching of English, 29, 1995, 37-58.
© National Council of Teachers of English, 1995
Abstract | Introduction | Development of the LRQ | Method | Results and Discussion | Discussion | Authors' Notes | References | Appendix 2: LRQ Items
A newly developed instrument, the Literary Response Questionnaire (LRQ), provides scales that measure seven different aspects of readers' orientation toward literary texts: Insight, Empathy, Imagery Vividness, Leisure Escape, Concern with Author, Story-Driven Reading, and Rejection of Literary Values. The present report presents evidence that each of these scales possesses satisfactory internal consistency, retest reliability, and factorial validity. Also, a series of five studies provided preliminary evidence that each scale may be located in a theoretically plausible network of relations with certain global personality traits (e.g., Absorption), with aspects of cognitive style (e.g., Regression in Service of the Ego), and with some of the learning skills that are relevant to effective work in the classroom (e.g., Elaborative Processing). In a variety of reaching and research settings, the LRQ may be a useful measure of individual differences in readers' orientation toward literary texts.
In the last two decades, reader response theorists have reconceptualized how readers engage literary texts, and such reconceptualizations have prompted teachers of literature to rethink classroom practices (Miall, 1993). However, persistent controversies have led at least one recent commentator to suggest that reader response theories have a past rather than a future (Freund, 1987, p. 10). What readers actually do, and what their activities imply about the status of literary texts remain very contentious topics. In fact, the principal theoretical statements in the area seem to suggest that readers of literary texts undertake only those activities that coincide with the tenets of the theorist's viewpoint. Thus, Iser's (1978) readers negotiate meaning in relation to the implied reader structured into the text; Fish's (1980) readers enact the modes of response authorized by their interpretive communities; Holland's (1968) readers search for identity themes in various narrative forms.
But studies in which readers think aloud as they read (Kintgen, 1983; Smith, 1991) have portrayed focal reader activities, such as the paraphrasing, thematizing, allegorizing, and problem-solving (Dias and Hayhoe, 1987) that do not neatly fit theoretical expectations. These studies suggest that readers are sufficiently self-aware to describe their own reading activities, perhaps not with critical or psychological precision (Hansson, 1990), but certainly with sufficient clarity to extend our understanding of their diverse approaches. If so, readers also may be sufficiently self-aware to provide valid descriptions of stable individual differences in reader response, that is, those differences among readers that persist despite variations in text (e.g., poetry versus prose) or circumstances (e.g., livingroom versus classroom). Although there is widespread appreciation of the importance of individual differences in reader response, there is currently available no sufficiently general and psychometrically satisfactory questionnaire for assessing such differences.
Therefore, it is appropriate to consider the range of beliefs, attitudes, and predilections that readers report and to develop a psychometrically sound means for assessing such variations. To this end, we have developed the Literary Response Questionnaire (LRQ). Its current version offers a set of low-inference scales that measure different aspects of readers' orientation toward literary texts. The purpose of the present report is to describe development of the LRQ, to review evidence of its reliability and validity, and to offer suggestions for its use in teaching or research settings.
Development of the Literary Response Questionnaire
Only a handful of questionnaires have been designed to examine approaches to reading literature, and several of these were intended for use with readers at the elementary or junior high school levels. In contrast, we developed the LRQ to assess variation among readers with a relatively well developed conception of literature. Among readers with at least an upper level high school background, would some of these readers typically prefer reading in order to escape awareness of their daily concerns (Nell, 1988)? Would some of them typically approach fiction in order to pursue the unfolding story-line (Hunt & Vipond, 1985)? Might some readers report that they read to gain insights into their own and others' feelings (Gold, 1990)? And would some indicate that they try to characterize the author's distinctive style and recurrent themes (Hirsch, 1967)? We thought it likely that readers would use a combination of such approaches, perhaps revealing themselves as readers through a distinctive profile of predispositions. Consequently the LRQ seeks to assess several significant aspects of readers' approaches to literary texts.
Our attempt to develop an instrument that would reflect different aspects of reader beliefs, attitudes, and predilections included attention to creating a tool that would not prejudge level of sophistication for the aspects identified. There may well be modes of literary response that are more "competent" than others (Culler, 1980), but current developments in literary theory encourage a cautious stance toward presuming the nature of a deep, a sensitive, or a competent reading. Contrasting approaches to reading, at least as measured by the LRQ, may have different but equally valued consequences (e.g., appreciation of the author's style, recognition of personal feelings), and our objective is to facilitate understanding of such variety rather than to identify what might constitute the most competent approach to literature. Thus, although the LRQ scales (and profiles ) eventually may be related to some standards of reading competence, we are not suggesting that as a primary objective.
Origins of LRQ Items
The LRQ went through several revisions. We wrote over 140 items, and we subjected these items to successive cycles of administration, evaluation, and revision. In preliminary versions the following aspects of reader activities were (at least temporarily) represented: empathizing with story characters; vividly imagining story settings, characters, or actions; perceiving correspondences with the reader's life-world; exploring the reader's personal identity; focusing on plot or story-line; attending to literary techniques (mainly diction); attempting to understand the author; perceiving cultural or social effects of literature; deriving ethical or moral viewpoints; reading for diversion or escape; making comparisons with other media; pursuing educational objectives; evaluating the "realism" of literature (or the lack thereof); creating a congenial reading environment. We identified this rather broad domain by consulting numerous studies of literary response. To the extent possible, we used the results of previous studies to frame specific items for the LRQ.
Previously developed questionnaires suggested several aspects of reader response. One frequently used instrument is Purves's Literary Transfer and Interest Measure (1973). Half of its 20 items assess Transfer (i.e., whether the reader relates literature to familiar life circumstances), and the other half assess Interest (i.e., whether the person reads frequently and is motivated to read by, for example, seeing a related movie). Purves reported correlations between Interest and reading for pleasure. Hynds (1985) found that for people high in Interest, complexity in judging people correlated with complexity in judging fictional characters. And, Miall (1987) found Transfer to be correlated with a separate measure of learning style: specifically, the self-reported tendency to elaboratively process study material (Schmeck, 1983). Due to these encouraging findings, we adapted some items from Purves' Transfer and Interest Measure when preparing items for the LRQ.
Another instrument developed by Purves (1973), the Response Preference Measure, consists of 20 items that -- demonstrated by Zaharias and Mertz (1983) -- assess 4 components of reader response: personal statements, descriptive response, interpretative response, and evaluative response. Using these scales, Zaharias (1986) observed differences in reader response across variations in text genre (fiction versus poetry) and tone (serious versus lighthearted). For example, readers were more likely to endorse personal statements about stories than about poems, and more likely to endorse descriptive responses to lighthearted literature than to serious literature. Although Zaharias's interest is in the influence of text variations on reader response, her results prompted the development of LRQ items that would assess analogous individual differences.
Questionnaires that are less familiar than those developed by Purves suggested additional LRQ items. For example, we adapted items from several short instruments devised by Tobin (1986) to assess whether students reacted to various literary techniques; whether they read for information, pleasure, or escape; and whether reading influenced their attitudes or behavior. Additionally, a number of items from a general reading questionnaire devised by Allerup (1985) suggested adaptability to reading literary texts as well (e.g., reading to avoid boredom, reading to improve reading skills). Saskia Tellegen has developed a range of questions in her work on reading with children, involving empathic reading, imaginal vividness, and reading for escape. Two items for the LRQ were suggested by her work (Tellegen & Coppejans, 1991). And we created items based upon responses to open-ended questions reported in several investigations, including: 1) Jacobsen's (1982) study of "literary space," asking about changes in sense of self during reading and feelings of creativity while reading; 2) Dickerson's (1988), study of personal reactions to literature, including questions about similarities between the reader and fictional characters and about recognizing one's own emotions in a text; and 3) Moffitt's (1987) study of readers of romance novels, with its questions about whether these readers' purpose in reading -- to escape their daily lives, to vicariously obtain cultural experiences -- might be relevant to the reading of traditional canonical texts.
Still other items were suggested by: 1) Koziol's (1982) questions concerning reading and culture, originally designed for teachers of literature; 2) Hunt and Vipond's (1985) description of story-driven approaches to literature; 3) the Denis (1982), the Sadoski, Goetz, and Kangiser (1988), and the Sadoski, Goetz, Clivarez, Lee, and Roberts (1990) studies of the role of imagery; 4) Miall's (1989; 1990) and Sadoski et al.'s (1988) studies of affective aspects of response; 5) Dias's work on readers' strategies (Dias & Hayhoe, 1987); and 6) Kuiken and his colleagues' phenomenological studies of responses to dreams and art (1989; 1993). Combined with the sources reviewed above, these ensured rather broad characterization of reader response in items devised for preliminary versions of the LRQ.
After four cycles of revision and assessment, the current version of the LRQ consists of 68 items, all positively worded. These items are rated for the extent to which "the statement is true of you" (1 = "not at all true" to 5 = "extremely true"). We used results from these items to determine the psychometric properties of the LRQ.
Subjects
The overall sample during development included 793 students at the University of Alberta who completed the current version of the LRQ. Administration at different times to rather differently constituted subsamples allowed replicated assessment of its psychometric properties. In one subsample, 407 Introductory Psychology students (239 women, mean age 20.3 years; 168 men, mean age 20.5 years) completed the LRQ during class time for course credit. In a second subsample, 275 Introductory Psychology students (171 women, mean age 22.3 years; 104 men, mean age 20.8 years) participated during class time for course credit. And 111 advanced undergraduate English students (59 women, mean age 27.8 years; 52 men, mean age 24.9 years) completed the LRQ while acting as paid participants in related experiments on reader response.
Basic Dimensions of the LRQ
To determine its dimensional structure, we analyzed responses to the LRQ using factor analysis. Factor analysis is a multivariate technique that minimizes the number of dimensions retained while simultaneously maximizing the informativeness of those dimensions. Factor analyses of the 68 items comprising the current version of the LRQ provided seven factors. (See Appendix 1 for a more detailed description of the factor analytic procedures used; Appendix 2 for the items that uniquely identify each factor, together with their primary factor loadings.) The meanings expressed by these seven factors can be summarized as follows:
1. Insight (14 items): This factor reflects an approach to reading in which the literary text guides recognition of previously unrecognized qualities, usually in the reader, but also in the reader's world. As indicated in Appendix 2, 9 items refer to shifts in self-understanding and 5 refer to changes in the reader's understanding of less personal matters.
2. Empathy (7 items): This factor indicates projective identification with fictional characters. Some items reflect the extended "presence" of these characters (e.g., in imagined dialogue), as though projective identification is regarded as a means to make the characters seem "real" to the reader.
3. Imagery Vividness (9 items): This factor expresses imaginary elaboration of a literary world that becomes vividly present not only visually, but also in feeling, sound, and smell.
4. Leisure Escape (11 items): This factor indicates an approach to reading that emphasizes reading for pleasure and as an enjoyable and absorbing departure from everyday responsibilities.
5. Concern with Author (10 items): This factor reflects interest in the author's distinctive perspective, themes, and style, as well as the author's biographical place in a literary or intellectual tradition.
6. Story-Driven Reading (8 items): This factor reflects an approach where the reader is focused on plot or story-line, with particular emphasis on interesting action and compelling conclusions.
7. Rejecting Literary Values (9 items): This factor represents the rejection of careful reading, of scholarly study, and of instructional presentation of literary texts. Reading literature is regarded as a compulsory and irrelevant task.
Table 1 summarizes evidence that 1) each of the 7 factors is replicable across subsamples; 2) there is close correspondence between each factor and a scale created by summing items that uniquely identifies that factor, and 3) each scale is internally consistent and possesses satisfactory test-retest reliability. (Appendix 1 describes the psychometric procedures used to substantiate these conclusions.)
Table 1. Psychometric properties of LRQ factors and scales __________________________________________________________________ Factor Factor x Alpha Test/retest Replication Scale R2 Coefficient Correlations Correlations __________________________________________________________________ Insight .979, .994 .89 .91 .75 Empathy .987, .990 .87 .85 .79 Imagery .989, .991 .86 .86 .78 Vividness Leisure .996, .995 .92 .92 .90 Escape Concern with .980, .992 .88 .86 .79 Author Story-driven .990, .993 .95 .81 .65 Reading Rejection of .995, .989 .90 .79 .75 Literary Values ___________________________________________________________________Superordinate DimensionsBecause results of the factor analysis indicated significant correlations between some factors, we conducted a second order factor analysis (Principal Components, Varimax rotation), revealing 2 distinct superordinate factors. As indicated in Table 2, on one factor, there were high values for Insight, Empathy, Imagery Vividness, and Leisure Escape. On the second factor, high values appeared for Story-Driven Reading and Rejection of Literary Values. Concern with Author split between these two superordinate factors. Thus, there is evidence that one second order factor collectively captures the engaging (Leisure Escape), perceptually replete (Imagery Vividness), and self-implicating (Empathy) modifications of meaning (Insight) warranting the label Experiencing (cf., Dewey 1934). And, there is evidence that another second order factor collectively captures the search for compelling narrative coherence (Story-Driven Reading). and inattention to literary complexity (Rejection of Literary Values) warranting the label Literal Comprehension. But, it should be emphasized that the first order factors are sufficiently independent to require separate examination in studies of reader response.
Table 2. Second Order LRQ Factors and Factor Loadings (>.400) _________________________________________________________________ Experiencing Literal Comprehension Insight .745 -- Empathy .775 -- Imagery Vividness .869 -- Leisure Escape .677 -- Concern with Author .420 -.676 Story-Driven Reading -- .859 Rejection of Literary Values -- .706 __________________________________________________________________Construct Validity of LRQ Scales
Observations bearing on the validity of the LRQ scales are available from a series of five studies undertaken simultaneously with development of the LRQ. Because of their interest, observations involving two early versions of the LRQ will be presented (Studies 1 and 2), but only for Insight and Leisure Escape, the two scales which remained essentially the same as their earlier counterparts in the final version of the LRQ. Likewise, the observations only obtain when comparable results occurred in both Studies 1 and 2. Primarily, observations based upon the final two (nearly identical) versions of the LRQ will be presented.
Method
The Studies
Study 1: We administered a 55-item version of the LRQ to 352 Introductory Psychology students. Of these participants, 153 also completed the Experience Inventory (Costa & McCrae, 1978), a measure of openness to experience, and the Sensitivity Questionnaire, a measure of aesthetic sensitivity (Child, 1965). Also, 77 participants completed a questionnaire concerning personally significant dream experiences (Kuiken & Sikora, 1993).
Study 2: We administered a 62-item version of the LRQ to 315 Introductory Psychology students and 75 upper level English students. Of these Study 2 participants, 88 (Psychology students only) also completed the Experience Inventory (Costa & McCrae, 1978) and the Sensitivity Questionnaire (Child, 1965), and 210 (Psychology students only) also completed the questionnaire concerning personally significant dream experiences (Kuiken & Sikora, 1993).
Study 3: Here, we administered. a 132-item version of the LRQ to 487 Introductory Psychology students and 61 upper level English students. (This version included 64 of the 68 items of the current LRQ, with minor working differences.) Of these participants, 470 also completed the Absorption Scale (Tellegen & Atkinson, 1974), a measure of openness to self-altering imaginal experiences, and 61 (all English students) completed the Inventory of Learning Processes (Schmeck, 1983), a measure of different learning styles.
Study 4: For this study, we administered a 96-item version of the LRQ to 407 Introductory Psychology Students. (This version included all 68 items of the current LRQ.) Of these participants, 260 also completed the Current Reading Questionnaire, a brief survey of non-curricular reading patterns developed by the authors; this questionnaire invited participants to rank order a range of leisure activities (such as seeing movies, participating in sports, and listening to popular music) and to indicate how often they read both literary and non-literary texts.
Study 5: Finally, we administered the same 96-item version of the LRQ to 275 Introductory Psychology students and 111 advanced undergraduate English students. Of these, 270 (Psychology students only) also completed the Absorption Scale, and 60 (English students only) also completed the Multidimensional Personality Questionnaire (Tellegen, 1982), a 300 item questionnaire that subsumes the Absorption Scale and 10 other factorially independent personality scales.
Results and Discussion
In the results reported below, * = p < .05, ** = p < .01, and *** = p < .001, all two-tailed.
Second Order Factors
Table 3. Construct Validation of LRQ scales _______________________________________________________________________________ Absorption Elab Read English/Psych St 3, St 4, St 5 Processes Novels Differences _______________________________________________________________________________ Insight .44c, .49c, .33c .43c .30c 3.53/2.90c Empathy .47c, .50c, .43c .39c .22c 2.68/2.16c Imagery Vividness .52c, .54c, .43c .39c .32c 3.82/3.24c Leisure Escape .37c, .30c, .15c .26c .58b 3.85/3.15c Concern with Author .26c, .34c, .19c .40c .31c 2.81/2.12c Story-driven Reading .05, -.05, -.07c -.17c -.20c 3.27/3.71c Rejection of Literary Values -.09, -.20c,-.12c -.27c -.36c 1.63/2.25c _______________________________________________________________________________ a = p < .05, b = p < .01, and c = p < .001, two-tailedAs indicated in Table 3, some variables from other scales correlated reliably with several LRQ scales in a pattern that lends credibility to the second order factors (Experiencing and Literal Comprehension) described above. First, in Studies 3, 4, and 5, the Absorption Scale consistently correlated with Insight, Empathy, and Imagery Vividness -- and more modestly with Leisure Escape and Concern with Author. Recall that these scales all had high results on the second order Experiencing factor. Because the Absorption Scale reflects readiness to be captured by imaginal events (e.g., "I can imagine things so vividly that they hold my attention as a good movie or story does") and readiness to modify them (e.g., "I can imagine that my body is so heavy that I could not move if I wanted to"), the LRQ scales on this superordinate factor may jointly reflect the absorbing elaboration of meanings that can occur while reading literary texts (e.g., recognizing previously overlooked feelings, empathically enlivening fictional characters, imaginally concretizing story scenes, comparing one of the author's themes with another).
This interpretation is substantiated by the observation that, in Study 3, each of the scales from the second order Experiencing factor positively correlated with the Elaborative Processing subscale of the Inventory of Learning Processes (e.g., "I learn new words or ideas by visualizing a situation in which they could occur"), a scale that Schmeck (1983) described as reflective of personalized elaboration of learning materials. Because these LRQ scales reflect such readiness to elaboratively respond to literary texts, it is not surprising that each of them predicted how frequently participants read novels (as reported on the Current Reading Questionnaire in Study 4) and that English students scored higher on each of these scales (Studies 3 and 5; see Table 3).
On the other hand, the two LRQ scales (Story-Driven Reading, Rejection of Literary Values) that distinguish the second order Literal Comprehension factor proved: 1) not reliably related (Story-Driven Reading) or inversely related (Rejection of Literary Values) to Absorption (Studies 3, 4, and 5); 2) inversely related to Elaborative Processing (Study 3); 3) inversely related to how frequently participants read novels (Study 4); and 4) more characteristic of Psychology students than English students (Studies 3 and 5; see Table 3). Besides confirming that the superordinate Literal Comprehension factor reflects low levels of interest in distinctly literary texts, this pattern also suggests that the modes of response expressed by the Story-Driven Reading and the Rejection of Literary Values scales lack the elaborative personalization associated with the Experiencing factor.
First Order Factors
Although validating evidence for the second order factors also contributes to the meaningfulness of the subordinate first order factors, the validity of the first order factors is more clearly indicated by correlations that are distinctive for each LRQ scale.
Insight. In Studies 1 and 2, the Insight scale (but not the Leisure Escape scale) reliably correlated with the Regression-in-the-Service-of-the-Ego subscale of the Sensitivity Questionnaire (e.g., "Unusual but unimportant aspects of a situation often intrigue me, occupying my attention and imagination"). For Study 1, the correlation is r = .30***; for Study 2, r = .30**. Following Kris, Child (1965) described this subscale as a measure of the ability to integrate regressive fantasy with mature thought. That such regressive/integrative thought is characteristic of insight-oriented reading is further suggested by the finding that the Insight scale (but again not Leisure Escape) correlated with a scale reflecting personal insights following dreaming (e.g., "After a dream I often feel sensitive to aspects of reality that I typically ignore"). Here, the correlations found are: Study 1, r = .36**; Study 2, r = .28***.
Moreover, since Child (1965) found the Regression-in-the-Service-of-the-Ego subscale predictive of esthetic judgment in visual art, it is consistent that the Insight scale correlated with the Esthetics subscale of the Experience Inventory (e.g., "I have had experiences that inspired me to write a poem or story"). For Study 1, the result is r = .26**; for Study 2, r = .34**.
In sum, the Insight scale may reflect a form of regressive/integrative thought during reading that subserves these readers' esthetic interests. Also, contrasting patterns of correlations substantiate the distinction between a form of absorbing reading that heightens awareness (Insight) and a form of absorbing reading that dulls awareness (Leisure Escape), a potentially important distinction anticipated by Nell (1988, p. 232) and (with some unnecessary psychoanalytic encumbrances) by Holland (1968, pp. 66, 92).
Empathy. In Study 5, Empathy correlated with the Stress scale of the Multidimensional Personality Questionnaire (e.g., "I sometimes get myself into a state of tension and turmoil when I think of the day's events; r = .21*) and with the Alienation scale of that instrument (e.g., "Most people make friends because they expect friends to be useful;" r = .26*). Thus, there is some evidence that negative affect, either anxious distress or estrangement, may be associated with the impulse to project oneself into the feeling-rich aspects of literary texts.
Imagery Vividness. Also in Study 5, Imagery Vividness correlated (r = .32**) with a scale measuring response inconsistency due to socially desirable responding on the Multidimensional Personality Questionnaire (MPQ). Although Imagery Vividness is the only LRQ scale to be correlated with any of the validity scales of the MPQ, this finding does indicate that, when reporting in a classroom setting, it may be socially desirable to report vivid imagery in response to literary texts.
Leisure Escape. In Studies 1 and 2, the Leisure Escape scale (but not the Insight scale) correlated with 1) the Feeling subscale of the Experience Inventory (e.g., "Feelings and emotions are important guides to conduct for me." For Study 1, r = .21*; for Study 2, r = .23*); 2) the Values subscale of the Experience Inventory (e.g., "The different ideas of right or wrong that people have in other societies may be right for them." Study 1, r = .28***; Study 2, r = .32**); and 3) the Tolerance for Complexity subscale of the Sensitivity Questionnaire (e.g., "Insofar as philosophy makes one doubt his basic beliefs, it should be encouraged;" for Study 1, r = .35***; for Study 2, r = .23*). On the one hand, these findings suggest that Leisure Escape is associated with openness, especially openness to complexity of feeling, but close examination of these items also suggests a somewhat assertive openness "ideology." This interpretation is supported by correlations between Leisure Escape and the Independence of Judgment subscale of the Sensitivity Questionnaire (Study 1: r = .36***; Study 2: r = .32**).
It may be noted that Leisure escape was the only LRQ scale to yield gender differences. In Study 5, women were significantly more likely to report Leisure Escape activities than were men (3.54 versus 3.11, p < .001).
Concern with Author. In general, the Concern with Author scale associated with interest in the fine arts beyond literature. On the Current Reading Questionnaire (Study 4), people who had scored high on Concern with Author reported that they more frequently read novels (r = .31**), read poetry (r = .27**), and listened to classical music (r = .22**). Also, of all the LRQ scales, Concern with Author correlated most highly with the Methodical Study subscale from the Inventory of Learning Processes (e.g., "I review course material periodically during the term." Study 3: r = .45**). Given these hints of disciplined attempts to understand the arts, it is noteworthy that people high on the Concern with Author scale also tended to be high on the Achievement scale from the Multidimensional Personality Questionnaire (e.g., "I often keep working on a problem even if I am very tired." Study 5: r = .19, p < .07).
Story-driven Reading. In Study 5, Story-Driven Reading correlated with the Tradition scale of the Multidimensional Personality Questionnaire (e.g., "I very much dislike it when someone breaks accepted rules of good conduct"; r = .34***) and with the Social Potency scale of that same instrument (e.g., "When I work with others, I like to take charge"; r =.21*). Also, in Study 3, Story-Driven Reading inversely related to the Methodical Study scale of the Inventory of Learning Processes (e.g., "I maintain a regular schedule of study hours"; r = .31**). And, in Study 4, as indicated by the Current Reading Questionnaire, Story-Driven Reading associated with more frequent movie-going (r = .21**) and TV watching (r = .29**). Perhaps Story-Driven Reading is associated with a decisive (even anti-intellectual) commitment to traditional values-and particular attention to the moral implications of a story line.
Rejection of Literary Values. In Study 5, Rejection of Literary Values associated with low scores on the Achievement scale of the Multidimensional Personality Questionnaire (r = -.23*) and high scores on the Aggression scale of that instrument (r = .26**). Also, in Study 3, Rejection of Literary Values was inversely related to the Methodical Study scale of the Inventory of Learning Processes (r = -.38**). And, as indicated by the Current Reading Questionnaire (Study 4), Rejection of Literary Values associated with frequent listening to popular music (Study 4: r = .19**), with frequent TV-watching (r = .22**), and with involvement in sports (r = .23**). Apparently Rejection of Literary Values involves aggressive resistance to careful reading of literary texts.
This interpretation is consistent with results from Studies 3 and 5 in which English students were asked to read a short story that had previously been examined for segments that were highly foregrounded and that, therefore, manifested stylistic variations at the phonetic, grammatical, and semantic levels (e.g., alliteration, ellipsis, metaphor). Rejection of Literary Values (but not Story-Driven Reading) inversely related (r = -.23**) to the tendency to spend more time on highly foregrounded passages (Miall & Kuiken, 1994). Thus, readers scoring high on this factor were relatively inattentive to the stylistic variations that are distinctly literary.
Discussion
Examination of their psychometric properties indicated that each of the LRQ scales proved to have very good internal consistency, retest reliability, and factorial validity. And, the sets of items comprising each scale are readily interpreted as a distinct aspect of readers' approach to literary texts. The series of studies described above affirmed the psychological meaningfulness of these scales by demonstrating that they may be located in a theoretically plausible network of relations with certain global personality traits (e.g., Absorption), with aspects of cognitive style (e.g., Regression in the Service of the Ego), and with some of the learning skills that are relevant to effective work in the classroom (e.g., Elaborative Processing). Because most of these measures have been independently validated by other investigators, the overall pattern of results provides promising evidence of the construct validity of the LRQ scales.
Although the LRQ is comprised of statements expressive of beliefs, attitudes, predilections, and behavior descriptions, there is no reason to take these statements simply at face value. These diverse statements do have face validity as measures of reader self-perceptions, and they do form several coherent and readily interpreted factors. But, in-depth understanding of these factors may prove complex. For example, self-reported preferences may not only reflect evaluations in actual reading situations; they may indirectly reflect the skills that enable reading in the preferred manner. And, self-reported behavior descriptions may likewise reflect not only what readers actually do while reading; they may indirectly reflect reader values or motives. Only future studies will clarify whether the LRQ simply predicts independent reader self-reports (e.g., what readers say in think-aloud studies), or whether it predicts methodologically diverse measures of reader skills and abilities (e.g., how readers perform in studies of reading comprehension).
Thus, although the LRQ was not devised as a test of literary competence, it is noteworthy that, compared with Psychology students, senior English students scored significantly higher on Insight, Empathy, Imagery Vividness, and Concern with Author, but lower on Story-Driven Reading. Assuming that advanced English students are relatively competent readers (both by virtue of self-selection and training), one interpretation is that this subset of LRQ scales indirectly reflects some of the skills by which readers make literary texts accessible, including the projective skills that enable empathic reading and the skills that make texts personally meaningful, including the elaborative skills that enable personal insights. While we have collected no independent evidence that would confirm a skill-based interpretation (e.g., performance on perspective-taking tasks), closer examination of the cognitive and affective capacities that are associated with LRQ scales is warranted.
On the other hand, some LRQ scales more plausibly reflect reader values (e.g., Rejection of Literary Values) and motives (e.g., Leisure Escape) than reader competence. Correlations between LRQ scales and certain personality dimensions (e.g., the Tradition and Achievement scales of the Multidimensional Personality Questionnaire) allude to socio-cultural constraints on the approaches to literary texts that are identified by the LRQ. Research relating developmental, educational, and social backgrounds to LRQ scales might give empirical credence to claims that the interpretive communities from which readers emerge constrain their approaches to reading.
Whether as skills or as values and motives, the LRQ scales suggest directions for the empirically grounded analysis of two global approaches to reading literature: Experiencing and Literal Comprehension. In the language provided by Cummins (1983), our analysis of the LRQ may be used to initiate a "functional decomposition" of these global approaches to reading. A functional decomposition begins by identifying components of a more inclusive mental construct and proceeds by determining how the functional relationships among those components contribute to the whole. Thus, functional decomposition of the Experiencing dimension might begin by acknowledging that the first order factors (e.g., Empathy, Imagery Vividness) are components of that more inclusive mental construct. As a next step, it may be useful to consider how those first order components jointly function to bring about Experiencing. For example, Imagery Vividness and Empathy may interact by providing concrete identification with literary characters; such concrete identification, in turn, may enable personal insight during reading. As another example, Empathy and Concern with Author may interact by accentuating the perspective of the author; such accentuation, in turn, may enable insights compatible with the author's perceived intentions. In brief, the functional decomposition of Experiencing and Literal Comprehension in terms of their respective first order factors suggests further directions for theory development and research.
One advantage of a decomposition based upon the LRQ is that relations between the superordinate factors and their first order components were determined using well-established psychometric procedures. The advantage of this method may be seen by comparing Experiencing and Literal Comprehension with Rosenblatt's (1978, 1986) conceptualization of Aesthetic and Efferent approaches to reading. Although Rosenblatt referred to reading events rather than individual differences, the LRQ Experiencing factor and her Aesthetic stance both involve empathic personal involvement, attention to vividly imagined narrative elements, and reflection on the life-world implications of the reading experience. And, the LRQ Literal Comprehension factor and her Efferent stance both involve focus on consensual text information such as literally paraphrasable meanings and directly designated narrative events. However, the Aesthetic and Efferent stances purportedly reflect opposite poles of a single continuum (cf. Many, 1991), whereas factor analysis of the LRQ indicated that Experiencing and Literal Comprehension are two factorially independent dimensions. Moreover, the Efferent stance purportedly involves consideration of an author's technique and socio-historical circumstances, whereas Concern with Author negatively related on the seemingly analogous Literal Comprehension factor. Such conceptual differences -- and their theoretical implications -- are difficult to resolve because, to date, the proposed components of the Aesthetic and Efferent stances have been articulated using quantitative procedures that provide "emergent" categories (cf. Hancock, 1993; Many, 1991) -- but not systematic psychometric information of the type reported here for the LRQ.
A second advantage to functional decomposition of the Experiencing and Literal Comprehension factors is that it may guide consideration of phenomena that occur at different levels of analysis. For example, a recent study by Many and Wiseman (1992) found that variations in teaching strategy did not reliably influence adoption of the Aesthetic stance in toto. Rather teaching strategy influenced adoption of particular components of that stance, for example imaging. By analogy, instructional encouragement of components of Experiencing such as Imagery Vividness may not facilitate the more complicated interactions among those components that identify Experiencing in toto. Investigators should be aware that sometimes the second order factors, and at other times the first order factors, of the LRQ may be implicated in the phenomena that they observe.
Finally, examination of discrete profiles of LRQ factors may facilitate examination of the needs of different types of readers. For example, Hunt and Vipond (1985) have suggested that it is possible to experimentally shift a reader from a story-driven to a point-driven approach to reading. But, are readers uniformly malleable? Perhaps Story-Driven Readers who also are capable of Vivid Imagery will be more readily influenced by such experimental manipulations than are readers who are Story Driven but lack the capacity for imaginatively concretizing textual meanings. Confirming this possibility would be one step toward clarification of instructional strategies that could effect long-term changes in approaches to reading.
There is, of course, no assurance that the LRQ will prove useful in research of the kind just suggested. Experience provides ample evidence that the utility of self-report measures cannot be taken for granted. Such measures have proven useful in some areas of research, but not in others, and they sometimes predict other self-report indices rather than the targeted behaviors. Moreover, they are subject to forms of bias, including socially desirable responding, that are difficult to control. Importantly, they do not necessarily capture the individual differences that truly matter in any particular area of study. Research beyond that reported here will be required to demonstrate that the LRQ will not fall victim to these limitations.
Alternatively, a prerequisite to the utility of any self-report instrument is the use of sound psychometric procedures during scale development. For example, the use of powerful multivariate techniques during revisions leading to the current form of the LRQ insures minimal inter-scale heterogeneity and inter-scale overlap. Similarly, our data indicate that the internal consistency and test-retest reliability of LRQ scales are more than adequate for research purposes. And our preliminary evidence indicates that (with the possible exception of Imagery Vividness) the promising construct validity of LRQ scales is not compromised by artifacts that commonly plague self-report measures, such as socially desirable responding. These aspects of scale development maximize the potential utility of the LRQ in a wide variety of relevant research paradigms, ranging from the observation of reading preferences and patterns in the classroom to the examination of reaction times and think-aloud protocols in laboratory studies of reader response. In the long term, knowledge of individual differences of the kind measured by the LRQ may enable teachers of literature to focus more productively on the needs of individual readers.
Authors' Notes: The central Research Fund of the University of Alberta and Program Grant No. 53-10018 from the Social Sciences and Humanities Research Council of Canada supported the research reported in this paper.
We gratefully acknowledge the assistance of Mathew Martin, Cam Balzer, and Mario Trono in the collection of data. We also thank Willie Van Peer for valuable comments on an earlier version of the LRQ and three anonymous reviewers for comments on an earlier version of this paper. We also gratefully acknowledge the assistance of Saskia Tellegen, who kindly made her work available to us in translation.
Software for computer administration of the LRQ (in both DOS and Windows formats) is available from the authors. Please send $20 US or $25 Canadian (checks made payable to David S. Miall) to cover costs. [See LRQ page]
We would welcome information from other researchers who employ the LRQ in their research efforts.
References
Allerup, P (1985). Why I like to read -- Statistical analysis of questionnaire data. Copenhagen, Denmark: Danish Institute for Educational Research.
Child, I. (1965). Personality correlates of esthetic judgment in college students. Journal of Personality, 33, 476-511.
Costa, P T., & McCrae, R. R. (1978). Objective personality assessment. In M. Storandt, I. C. Siegler, & M. F Elias (Eds.), The clinical psychology of aging (pp. 119-143). New York: Plenum Press.
Culler, J. (1980). Prolegomena to a theory of reading. In S. R. Suleiman & 1. Crosman (Eds.), The reader in the text: Essays on audience and interpretation (pp. 46-66). Princeton, NJ: Princeton University Press.
Cummins, R. (1983). The nature of psychological explanation. Cambridge, MA: MIT Press.
Denis, M. (1984). Imagery and prose: A critical review of research on adults and children. Text, 4, 381-401.
Dias, P, & Hayhoe, M. (1987). Developing response to poetry. Philadelphia, PA: Milton Keynes: Open University Press.
Dickerson, M. J. (1988, May). A voice of one's own: Creating writing identities. Paper presented at the Conference on College Composition and Communication, St. Louis, MO.
Dewey, J. (1934). Art as experience. New York: Capricorn Books.
Dollerup, C. (1985). The RAM studies 1, 2, 3, and 4: A Description of the thinking, set-up, and planning of the studies. University of Copenhagen, Copenhagen, Denmark: Research Unit for Experimental Studies in Reader Response.
Fagan, W. T., Cooper, C. R., & Jensen, J. M., Eds. (1975). Measures for research and evaluation in the English language arts. Urbana, IL: National Council of the Teachers of English.
Fish, S. (1980). Is there a text in this class? The authority of interpretive communities. Cambridge, MA: Harvard University.
Freund, E. (1987). The return of the reader. London: Methuen.
Gold, J. (1990). Read for your life: Literature as a life support system. Markham, Ontario, Canada: Fitzhenry & Whiteside.
Hancock, M. R. (1993). Exploring the meaning-making process through the content of literature response journals: A case study investigation. Research in the Teaching of English, 27, 335-368.
Hansson, G. (1990). Reading and understanding literature (Report Ser. No. 4.5). Albany, NY: State University of New York, Center for the Learning and Teaching of Literature.
Hirsch, E. D. (1967). Validity in interpretation. New Haven: Yale University Press. Holland, N. H, (1968). The dynamics of literary response. New York: Oxford University Press.
Hunt, R. A., & Vipon, D. (1985). Crash-testing a transactional model of literary reading. Reader: Essays in Reader-Oriented Theory, 14, 23-39.
Hynds, S. (1985), Interpersonal cognitive complexity and the literary response processes of adolescent readers. Research in the Teaching of English, 19, 386-402.
Iser, W. (1978). The act of reading: A theory of aesthetic response. Baltimore, MD: Johns Hopkins University Press.
Jacobsen, M. (1982). Looking for literary space: The willing suspension of disbelief re-visited. Research in the Teaching of English, 16, 21-38.
Kintgen, E. (1983). The perception of poetry. Bloomington, IN: Indiana University Press.
Koziol, S. (1982). Responding to literature. Communication skills. (PCRP Assessment Survey I). Harrisburg, PA: Pennsylvania State Department of Education.
Kuiken, D., Schopflocher, D., & Wild, T. C. (1989). Numerically aided methods in phenomenology: A demonstration. Journal of Mind and Behavior, 10, 373-392.
Kuiken, D., & Sikora, S. (1993). The impact of dreams on waking thoughts and feelings. In A. Moffit, R. Hoffman, & M. Kramer (Eds.), Functions of Dreams (pp. 419-476). Albany, NY: State University of New York Press.
Many, J. E. (1991). The effects of stance and age level on children's literary responses. Journal of Reading Behavior, 23, 61-85.
Many, J. E., & Wiseman, D. L. (1992). The effect of teaching approach on third grade students' response to literature. Journal of Reading Behavior, 24, 265-287.
Miall, D. S. (1987). Learning in autonomous student groups: Learning skills as a predictor of satisfaction. Studies in Educational Evaluation, 13, 175-183.
Miall, D. S. (1989). Beyond the schema given: Affective comprehension of literary narratives. Cognition and Emotion, 3, 55-78.
Miall, D. S. (1990). Readers' responses to narrative: Evaluating, relating, anticipating. Poetics, 19, 323-339.
Miall, D. S. (1993). Constructing understanding: Emotion and literary response. In D. Bogdan & S. B. Straw (Ed.), Constructive reading: Teaching beyond communication (pp. 61-81). Portsmouth, NH: Boynton/Cook.
Miall, D. S., & Kuiken, D. (1994) Foregrounding, defamiliarization, and affect: Response to literary stories. Poetics, 22, 389-407.
Moffitt, M. A. (1987, May). Understanding the appeal of the romance novel for the adolescent girl: A reader-response approach. Paper presented at the Meeting of the International Communication Association, Montreal, Canada.
Nell, V. (1988). Lost in a book. New Haven: Yale University Press.
Purves, A. (1973). Literature education in ten countries. An empirical study. New York: Wiley & Sons.
Rosenblatt, L. M. (1978). The reader, the text, the poem. Carbondale, IL: Southern Illinois University Press.
Rosenblatt, L. M. (1986). The aesthetic transaction. Journal of Aesthetic Education, 20,122-128.
Rozeboom, W W. (1991a). HYBALL: A method for subspace-constrained oblique factor rotation. Multivariate Behavioral Research, 26, 163-177.
Rozeboom, W. W. (1991b). Theory and practice of analytic hyperplane detection. Multivariate Behavioral Research, 26, 179-197.
Sadoski, M., Goetz, E. T., & Kangiser, S. (1988). Imagination in story response: Relationships between imagery, affect, and structural importance. Reading Research Quarterly, 23, 320-336.
Sadoski, M., Goetz, E. T., Olivarez, A., Lee, S., & Roberts, N. M. (1990). Imagination in story reading: The role of imagery, verbal recall, story analysis, and processing levels. Journal of Reading Behavior, 22, 55-90.
Schmeck, R. R. (1983). Learning styles of college students. In R. F Dillon & R. R. Schmeck (Eds.), Individual differences in cognition (pp. 233-279). New York: Academic Press.
Smith, M. W. (1991). Constructing meaning from text: An analysis of ninth-grade reader responses. Journal of Educational Research, 84, 262-271.
Tellegen, A. (1982). Brief manual for the multidimensional personality questionnaire. Unpublished manuscript. University of Minnesota, Department of Psychology, Minneapolis, MN.
Tellegen, A., & Atkinson, G. (1974). Openness to absorbing and self-altering experiences ("absorption"), trait related to hypnotic susceptibility. Journal of Abnormal Psychology, 83, 268-277.
Tellegen, S., & Coppejans, L. (1991). Verbeeldend lezen. The Hague: NBLC.
Tobin, A. R. (1986). A study of the effectiveness of literature instruction demonstrated in two samples of high school juniors. Unpublished master's thesis, Kean College of New Jersey, New Brunswick, NJ.
Zaharias, J. A. (1986). The effects of genre and tone on undergraduate students' preferred patterns of response to two short stories and poems. Research in the Teaching of English, 20, 56-68.
Zaharias, J. A., & Mertz, M. P (1983). Identifying and validating the constituents of literary response through a modification of the response preference measure. Research in the Teaching of English, 17, 231-241.
Appendix 1: Psychometric Properties
In the present study, a Principal Components factor analysis was followed by HYBALL (oblique) factor rotation (with SPIN) to identify optimal axis positions (Rozeboom, 1991x, 1991b). Analysis of the 68 items comprising the current version of the LRQ provided 7 factors accounting for 50.4% of the total variance in the overall sample. With the exception of one marginal case, an item was selected as expressive of a factor if 1) its loading on that factor after rotation was greater than.400; 2) all other loadings for that item were less than .400 and also less than 75% of the loading on the relevant factor; and 3) the preceding criteria were met not only in the analysis of the overall sample but also in factor analyses of each of the two subsamples. Items meeting these criteria are presented in Appendix 2, together with factor loadings derived from the overall sample.
Replicability of these 7 factors was assessed by examining correlations between the factor scores derived from factor analysis of the overall sample and the analogous factor scores derived from separate factor analyses of the two subsamples. These correlations (see Table 1) were very high for all factors, indicating close correspondence between the optimal estimate of these factor scores (the overall sample) and the factor scores derived from two somewhat divergent subsamples (Psychology students versus Psychology and English students).
Ratings for items selected as expressive of each factor (see above) were equally weighted and summed to create 7 LRQ scales. The squared correlations between these item composites and their respective factor scores (see Table 1) indicated that they captured from 83-93% of the variance in the target factors. These squared correlations were never more than .02 less than the squared multiple correlations between these items and the target factors scores. Thus, LRQ scales based on equally weighted sums reflect the original factors nearly as well as do optimally weighted sums.
As indicated in Table 1, all 7 scales show satisfactory to excellent internal consistencies, as indicated by alpha coefficients calculated on data from the overall sample. Furthermore, evidence of retest reliability was provided by 123 Introductory Psychology students who completed the LRQ a second time, ten weeks after initial administration. These reliability estimates generally substantiate the claim that the scales measure rather stable individual differences, although correlations do range from .65 for Story-Driven Reading to .90 for Leisure Escape.
Appendix 2: LRQ Items and Their Primary Factor Loadings
Insight
Reading literature makes me sensitive to aspects of my life that I usually ignore (786, personal).
In literature I sometimes recognize feelings that I have overlooked during my daily life (.775, personal).
I often find my shortcomings explored through characters in literary texts (.734, personal).
I find that literature helps me to understand the lives of people that differ from myself (.732, non-personal).
Reading literature often gives me insights into the nature of people and events in my world (.728, non-personal).
I often see similarities between events in literature and events in my own life (.723, personal).
I often find my own motives being explored through characters in literary texts (.715, personal).
I find that certain literary works help me to understand my more negative feelings (.711, personal).
Literature enables you to understand people that you'd probably disregard in normal life (.700, non-personal).
I sometimes find that reading a literary text makes me feel like changing the way I live (.625, personal).
In my reading, I learn to recognize more readily certain types of people or events, i.e., I can see these types more clearly after reading about a particular example in a literary text (.619, non-personal).
When I begin to understand a literary text, it's because I've been able to relate it to my own concerns about life (.602, personal).
Literature often gives special emphasis to those things that make a moral point (.513, non-personal).
Sometimes while reading literature my feelings draw me toward a distinctly unsettling view of life (.512, personal).Empathy
Sometimes I feel like I've almost "become" a character I've read about in fiction (.856).
I sometimes have imaginary dialogues with people in fiction (786).
When I read fiction I often think about myself as one of the people in the story (.737).
I sometimes wonder whether I have really experienced something or whether I have read about it in a book (677).
1 actively try to project myself into the role of fictional characters, almost as if I were preparing to act in a play (.652).
Sometimes characters in novels almost become like real people in my life (.647).
After reading a novel or story that I enjoyed, I continue to wonder about the characters almost as though they were real people (.509).Imagery Vividness
I often see the places in stories I read as clearly as if I were looking at a picture (.800).
I can readily visualize the persons and places described in a novel or short story (.723).
I sometimes think I could draw a map of the places I have read about in a work of fiction (.660).
Sometimes a scene from a story or poem is so clear that I know its smell, its touch, its "feel" (.638).
I often hear dialogue in a novel as though I were listening to an actual conversation (560).
When I read a literary text, a scene that is only partly described often becomes a whole, vividly present place in my mind (.545).
When reading a story, sometimes I can almost feel what it would be like to be there (.515).
I usually hear the tone of speech in a dialogue from a story or novel (498).
Often when I read literary texts, descriptions of smells suggest colors, descriptions of colors suggest feelings, and so on (.468).Leisure Escape
Sometimes I like to curl up with a good book just to enjoy myself (840). When I have spare time my favorite activity is reading a novel (.817).
Very often I cannot put down a story until I have finished reading it (.796).
Reading literature is a pleasurable way to spend time when I have nothing else to do (.774).
Reading a story is a wonderful way to relax. (.763).
While reading I completely forget what time it is (.740).
I find that reading literature is a great help in taking my mind off my own problems (666).
I like to become so absorbed in the world of the literary text that I forget my everyday concerns (.608).
Once I've discovered one work by an author I like, I usually try to read all the other works by that author (.579).
I am often so involved in what I am reading that I am no longer aware of myself (.578).
I often wish I had more time for reading literature (.509).Concern With Author
One of my primary interests in reading literature is to learn about the themes and concerns of a given author (.755).
In reading I like to focus on what is distinctive about the author's style (.742).
One of my primary interests in reading is to learn about the different genres of literature (.727).
I like to see how a particular author's work relates to other literature of the author's period (.726).
When reading I usually try to identify an author's distinctive themes (.701).
One of my primary interests in reading literature is to appreciate the author's understanding of society and culture (.686).
I think literature is especially interesting when it illuminates facts about the author's life (.610).
When I find a work of literature I like, I usually try to find out something about the author (.608).
The challenge of literature is to comprehend the author's unique view of life (.605).
I am often intrigued by an author's literary technique (.508).Story-Driven Reading
I like to see tension building up in the plot of a story (659).
The type of literature I like best tells an interesting story (.635).
I think the most important part of fiction or drama is plot (.619).
When reading a novel, what I most want to know is how the story turns out (.609).
I like it best when a story has an unexpected ending (.600).
I prefer to read fiction in which there is plenty of action (.599).
When reading a novel my main interest is seeing what happens to the characters (.576).
1 find it difficult to read a novel in which nothing much seems to happen (.540).Rejection of Literary Values
I think people should spend less time talking or writing about literature (.755).
Even if literature were well taught, I think high schools should not devote so much time to it (.738).
For me a work of literature is destroyed by trying to analyze it (.711).
One of the things I dislike most about being a student of literature is the teacher who tells you what a literary text means (.703).
Reading literary texts from past centuries should be left to literary scholars and historians (.623).
1 don't believe that literature is socially relevant (.616).
1 disliked English in high school because most of the texts I was asked to read I would not have chosen myself (.579).
Works of literature often seem to make the issues of life more complicated than they actually are (.491).
If I want to spend time reading, I don't choose "literary" texts (.392).
return to Reader Response Document created August 8th 2002