You are on page 1of 11

Journal £>f Educational Psychology Copyright 2000 by the American Psychological Association, Inc.

2000,\fol.92,No. 1,96-106 0022-0663/00/$5.00 DOT: 10.1037//0022-0663.92.1.96

Reader Beliefs and Meaning Construction in Narrative Text


Gregory Schraw
University of Nebraska—Lincoln

This research examined how transmission and transaction beliefs about reading affect
comprehension, engagement, and holistic understanding of narrative text. Transaction beliefs,
which emphasize reader-generated meaning, were related positively to the type and number of
reader responses as well as to the sophistication of one's holistic interpretation. Transmission
beliefs, which emphasize understanding the author's intended meaning, were unrelated to all
outcome measures. Transmission and transaction beliefs were unrelated to each other. These
findings support the view that transaction beliefs facilitate meaning construction of narrative
text, whereas transmission beliefs do not.

Skilled readers use a number of strategies to construct more critical perspective in which they actively questioned
meaning from text (Kintsch, 1998). These include lower and transformed text, whereas high school students inter-
order decoding skills such as lexical retrieval as well as ested in history read from a less critical perspective without
higher order comprehension skills such as identifying main seriously questioning the legitimacy of text assertions.
ideas and drawing inferences. Skilled readers also invoke a Wade, Thompson, and Watkins (1994) reported similar
variety of beliefs about the reading processes that affect their findings in which expert historians understood episodes
conceptual understanding of the text (Dole & Sinatra, 1994; from the Public Broadcasting System series The Civil War in
Garner & Alexander, 1994; Mathewson, 1994). These a much more sophisticated fashion than novices. Experts
include beliefs about reading self-efficacy (Schunk & Zim- invoked different beliefs about the events contained in the
merman, 1997; Shell, Murphy, & Bruning, 1995), the video compared with nonhistorians. Experts also tended to
credibility of the author's message (Dole & Sinatra, 3994), be more critical and presented more counterexplanations to
personal ideologies (Bogdan, 1990; Thomson, 1993; Wade, the theories discussed in the video.
Thompson, & Watkins, 1994), and beliefs about the meaning
construction process (Schraw & Bruning, 1996; Winehurg, Schraw and Bruning (1996) distinguished between trans-
1991). mission and transaction beliefs about reading. Transmission
The purpose of the present research is to explore how beliefs emphasize understanding the author's intended mean-
different types of beliefs affect meaning construction when ing; transaction beliefs emphasize the role of constructing
reading a highly interpretable narrative text. Beliefs about meaning on the basis of the reader's goals and purposes for
the reader's role in meaning construction may be especially reading. According to Bogdan and Straw (1990), the trans-
important when reading narrative texts because they are mission model views reading as a conduit in which informa-
more open to interpretation than are expository texts (Zwann, tion flows in a bottom-up and linear direction from the
1994). This study examines the contribution of transmission author to the reader (Hunt, 1990; Straw, 1990). Meaning is
and transaction beliefs to three aspects of reading comprehen- conveyed directly to the reader by means of the author's
sion, including understanding main ideas, generating per- ideas stated explicitly in the text. Transmission reading is
sonal responses to the text, and constructing a holistic reconstructive rather than constructive because meaning
interpretation of the text's meaning. flows directly from author to reader without changes in
meaning.
Transaction and Transmission Beliefs In contrast, the transaction model is based on the assump-
About Text Meaning tion that a text means different things to different readers
regardless of what the author intended. The transaction
Readers hold different beliefs about the meaning construc- model views reading as a dynamic system (Straw, 1990;
tion processes that affect their understanding (Alexander & Tompkins, 1980) or a reciprocal network among the reader,
Dochy, 1994; Chambliss, 1994; Straw & Sadowy, 1990). For text, and author (Rosenblatt, 1994). Information flows
example, Wineburg (1991) reported that professional histori- interactively to and from the reader. Meaning does not exist
ans differed from high school students on several dimen- in either the text or author per se but is constructed uniquely
sions of text understanding. Historians read texts from a by each reader, depending on his or her specific goals and
intentions. Transaction reading is constructive rather than
I thank Roger Bruning for his helpful comments on an earlier reconstructive because meaning is created by the reader
version of this article. (Harste, Burke, & Woodward, 1994).
Correspondence concerning this article should be addressed to
Gregory Schraw, Department of Educational Psychology, 309 The two aforementioned models suggest that transmission
Bancroft Hall, University of Nebraska, Lincoln, Nebraska 68588- and transaction beliefs make different contributions to
0345. Electronic mail may be sent to gschraw@unl.edu. reading. Transmission beliefs may increase memory for

96
READER BELIEFS 97
important information or salient events in the text given the meant"), and one that measures transaction beliefs (e.g., "I
reader's focus on the author's intended message. Transmis- like to interpret what I read in my own unique way"; see
sion beliefs may also promote the construction of an Appendix A). Interpretative responses were coded into 13
interpretation directly consistent with the text's main ideas. subcategories compiled from previous research reported by
In contrast, transaction beliefs may have little impact on Many and Cox (1992) and Schraw (1997). The 13 subcatego-
memory for factual information yet promote the construc- ries were subsumed within three broader categories, which
tion of a variety of interpretations, especially those that are described in greater detail in the Method section. The
diverge from the text's main ideas but are based on the thematic response category included statements about the
reader's own beliefs and expectations. Transaction beliefs themes and overall main ideas discussed in the story.
may also promote personal responses and affective engage- Critical responses included statements about the author's
ment compared with transmission beliefs (Farrell & Squire, intention or the text's clarity, such as whether the text was
1990; Rosenblatt, 1994). clear, well written, and achieved its goals. Personal re-
sponses included statements about personal reactions to the
story, such as interest, engagement, and the extent to which
Overview of the Present Research the reader related information in the story to his or her life.
The holistic interpretation was scored on a 9-level scale
The purpose of the present research is to examine how
(described in Schraw, 1997) that ranged from no interpreta-
transmission and transaction beliefs affect meaning construc-
tion to an explicit interpretation supported by events from
tion when reading a narrative text. Understanding the
the text and the reader's personal life.
contribution of each type of belief is important because there
currently are no empirical data that address how different
beliefs are related to reader responses and the construction
Research Questions
of a holistic text interpretation in a highly interpretable text.
Most previous research has focused on the role of inference- I focused on four research questions that examine the
making strategies in expository texts (Garner & Hansis, relationship among reader beliefs and several measures of
1994; Graesser, Singer, &Trabasso, 1994; Hartman, 1995). understanding:
Several studies have considered the relationship between 1. Is there a statistical relationship between the transmis-
prior knowledge and the construction of higher level text sion and transaction scales included in the RBI?
interpretations such as a situation model (Kintsch, 1998; 2. Are transmission and transaction beliefs related to
Zwann & Radvansky, 1998). Others have compared expert multiple-choice test performance?
and novice readers' constructive strategies when reading 3. Are transmission and transaction beliefs related to
historical texts (Wade etal., 1994; Wineburg, 1991). None of individual interpretative responses?
these studies used narrative text or explicitly controlled for 4. Are transmission and transaction beliefs related to
reader beliefs. holistic text interpretations?
The present research is the first to explicitly examine the My first question is whether there is a statistical relation-
relationship between beliefs about reading and the meaning ship between the transmission and transaction scales in-
construction process in a narrative text. A recent study by cluded in the RBI. This question provides a replication of the
Schraw and Bruning (1996) examined the relationship findings by Schraw and Bruning (1996) who reported no
between transmission and transaction beliefs and preposi- correlation between the two scales. Replicating this finding
tional recall in an expository text. Transaction beliefs were is important for both theoretical and practical reasons. From
related positively to recall, whereas transmission beliefs a theoretical perspective, a close replication indicates that
were related negatively to recall. The present study extends the instrument is reliable across samples and supports the
this finding by examining a variety of reader responses and view that transmission and transaction beliefs are indepen-
the extent to which readers construct a holistic interpretation dent (i.e., uncorrelated). Independent scales suggest that
of a narrative text. College students were asked to read a readers maintain separate transmission and transaction be-
1,000-word narrative text, complete a 20-item multiple- liefs and that each have a unique effect on reading. Thus,
choice test of main ideas, and write a two-page interpretative readers hold separate beliefs about the role of the author's
essay in which they described what they thought the story intended meaning that are independent of beliefs about the
meant and how it made them feel (see the Method section for reader's constructive role in meaning making.
further details). I selected a narrative text that was amenable From a practical perspective, separate scales allow re-
to a variety of interpretations to maximize the role of searchers to create four mutually exclusive categories con-
transaction beliefs in reading. An informationally dense sisting of high-transmission-high-transaction, high-transmis-
expository text may minimize the effect that transaction sion-low-transaction, low-transmission-high-transaction, and
beliefs have on reader response and higher order construc- low-transmission-low-trans action readers. Of special inter-
tive processes. est is whether high-transmission-high-transaction readers
Transmission and transaction beliefs were measured with differ from low-transmission-high-transaction readers. The
the 16-item Reader Belief Inventory (RBI) developed by former may adopt different reading strategies given that they
Schraw and Bruning (1996). This instrument includes two believe it is important to understand the author's intended
subscales, one of which measures transmission beliefs (e.g., message as well as their own constructed meaning. A
"When I read, I try to carry away exactly what the author comparison of multiple-choice test scores, interpretative
98 SCHRAW

responses, and holistic interpretation scores across these Materials


four possible groups enables one to isolate the unique
contribution of transmission and transaction beliefs (i.e., The target materials consisted of (a) the RBI, (b) an 870-word
their main effects) as well as their interaction. I predicted story titled The Book ofSand by Borges (1969/1977), (c) a 20-item
that transmission beliefs would be independent of transac- multiple-choice recognition test covering main ideas included in
the story, and (d) a response booklet in which each participant
tion beliefs, supporting a two-factor model. This finding
wrote a two-page reaction to the text. The RBI consists of a 16-item
would replicate Schraw and Bruning (1996). questionnaire developed by Schraw and Bruning (1996) in which
My second research question is whether transmission and individuals indicated on a 5-point Likert scale the degree to which
transaction beliefs are related to multiple-choice test perfor- they agreed or disagreed with each statement about the text (see
mance. There are no current studies that address this Appendix A). This instrument identified the relative strength of
question. Reader response theories (Rosenblatt, 1994) gener- transmission (e.g., "Good readers remember most of what they
ally predict that transaction beliefs should be related posi- read verbatim") and transaction (e.g., "I enjoy interpreting what I
tively to text understanding. Thus, one would expect transac- read in a personal way") beliefs.
tion beliefs to improve recognition of main ideas. However, The text consisted of the narrative story The Book of Sand
it also is possible that transmission beliefs improve recogni- (Borges, 1969/1977). Although the story contains an easy-to-
tion by focusing readers' attention on the text's main ideas. follow surface plot, it also contains a number of richly symbolic
interpretations beyond its explicit storyline. This story was selected
Thus, I expected both types of beliefs to be related positively
because it is brief yet compelling, invites a number of parallel
to recognition of main ideas. interpretations beyond its surface meaning, contains a great deal of
My third research question is whether transmission and imagery, and evokes a feeling of mystery and moral ambiguity that
transaction beliefs are related to individual interpretative was likely to elicit a wide variety of interpretative responses. (The
responses. I predicted that transaction beliefs are related Book of Sand is available from Gregory Schraw.)
positively to the thematic, critical, and personal responses, The recognition test included 20 four-option multiple-choice
whereas transmission beliefs are not related to these re- questions that address main ideas from the story (see Appendix B).
sponses. This prediction is in line with claims from the The 20 items used on this test were selected from a pool of 35 items
reader response literature (Farrell & Squire, 1990; Many & that were normed with 15 additional participants. All final test
items were answered correctly by no more than 90% and no less
Cox, 1992; Rosenblatt, 1994; Tompkins, 1980). Specifically,
than 50% of test takers. The overall success rate on the test was
I predicted that transaction beliefs would increase sophisti- roughly 75%. Coefficient alpha for the test was .86.
cated thematic responses such as generating hypotheses and Written responses were completed after the multiple-choice test.
thematic statements. In addition, I expected an increase in Individuals were given brief instructions asking them to describe
the total number and variety of personal responses such as what they thought the story meant and what kinds of personal
interest, affective engagement, empathy, and relating text thoughts and feelings it evoked in them (see Appendix C).
events to personal life. Responses were grouped into three response categories: thematic,
critical, and personal. Thirteen subcategories (5 under thematic, 3
My fourth research question is whether transmission and under critical, and 5 under personal responses) were identified in
transaction beliefs are related to holistic text interpretations. part on the basis of a content analysis of randomly selected
To answer this question, each response was assigned to one protocols (Weber, 1985) and similar categories proposed by Many
of nine mutually exclusive, interpretative categories. I and Wiseman (1992) and Schraw and Bruning (1996). The
predicted that transaction beliefs would be related positively thematic responses category includes retellings, text-based elabora-
to these nine levels of interpretation. This prediction is based tions, interpretation of symbols, generated hypotheses and explana-
on the assumption that transaction beliefs encourage readers tions, and thematic statements (see Appendix D). The critical
to construct a unique interpretation of the text by using main responses category includes statements pertaining to the author's
ideas found in the text along with their own prior knowledge purpose or intention, critical analysis of text ideas, and critical
and expectations (Farrell & Squire, 1990). Transacting with analysis of text structure. The personal responses category includes
statements pertaining to engagement and interest, cognitive reac-
text should lead to deeper processing as well as to a broader
tions, affective reactions, or events or characters in the text, and
synthesis of text themes (Rosenblatt, 1994). In contrast, statements relating events in the text to one's personal experiences.
transmission beliefs should focus one's attention on main
I (along with a trained research assistant) scored each written
ideas rather than constructing one's own interpretation of the protocol using the criteria in Appendix D. These protocols served
text. Presumably, transmission beliefs also limit the number twb purposes. One purpose was to quantify the type and number of
and type of personal responses to a text. reader responses. These responses were placed into one of the 13
mutually exclusive subcategories described previously. The two
judges scored each protocol together. Initial agreement was approxi-
Method mately 86%. All disputes were settled in conference by discussing
the response in the context of the entire essay. Approximately 1%
Participants of the statements were eliminated during scoring because of a lack
of agreement between the judges. Approximately 4% of the
Two hundred forty-seven undergraduates from a large midwest- comments were not scored because they were irrelevant to the
era university participated as part of their regular coursework. All essay, redundant, or rhetorical in nature.
participants were enrolled in sections of an introductory educa- The second purpose of the written protocols was to measure the
tional psychology class and participated outside of their regularly degree to which readers constructed a holistic text interpretation.
scheduled class in partial completion of a course research Criteria for this measure were adapted from Schraw (1997) who
requirement. developed a 9-level scale to assess overall interpretative meaning.
READER BEUEFS 99
The purpose of the holistic text interpretation score was to measure Level 7 represented an elaborated holistic model, which includes a
meaning construction beyond the level of individual main ideas comprehensive interpretation or comparison of one or more themes
and, in particular, whether readers possess an integrated understand- with detailed justification of those themes, as the following excerpt
ing of die text. This measure is important because previous studies illustrates:
have relied on objective measures of understanding, such as recall
and multiple-choice tests, that measure understanding at the This story shows how easy it is to become obsessed with
finding answers to things we don't understand. There are so
macropropositional text level rather than at the higher level of a many things that we can't understand that it is difficult to
situation model (Kintsch, 1998; Zwann, 1994). In the present case, admit not being able to figure them out. At certain times man
the holistic text interpretation score was used to measure one's will be defeated in his quest for knowledge. This doesn't mean
overall constructed interpretation of the text. we should give up easily, however. An example would be the
The nine levels of the holistic interpretation score are shown in search for answers about God. We can always try to I earn
Appendix E. Scores ranged from 0, which corresponded to no more about the existence of God, but we also know that there
are many things that we will never fully understand until the
holistic interpretation, to 8, which corresponded to an integrated end of time. But we must keep trying. Trying to learn more
interpretation in which the reader stated at least one main theme, about something that challenges us can only help, whether you
supported that theme in detail, and related it to his or her personal reach your final answer or not.
life.
Each protocol was assigned to one of nine mutually exclusive Level 8 consisted of an integrated holistic model, which relates an
levels of increasing sophistication. Level 0 consisted of nonthe- elaborated holistic model to personal experiences or philosophy, or
matic responses, such as repeating main ideas or information stated to another text. The following excerpt provides an example of
in the text, or comments related to writing the essay. None of the relating an interpretation to one's philosophy:
essays provided any information regarding an overall theme or text
interpretation, nor did they include personal reactions related to the It was obvious that The Book of Sand had already drained the
text or nonthematic elaborations of the text. Level 1 consisted of salesman's life and he wanted to get rid of it. The narrator
made the mistake of trading for it because he lost his Bible and
personal reactions and opinions without any text-based elabora- pension (i.e., a symbol of eternal life) for something that
tions, interpretation of symbols, or thematic inferences. These would ultimately consume him. I find a sharp contrast
essays essentially offered personal opinions about the text without between the Bible and The Book of Sand that is very
stating a main theme. Level 2 included text elaborations without important The Book of Sand makes no sense and is a heavy
clarification of symbols or themes. Text elaborations consisted of burden, and ultimately drives yoa insane. The Bible, on the
inferences that were based on information in the text {e.g., other hand, is eternal truth, has a first and a last page, has a
"Someone else will come along someday and find the book in the message of hope that relieves you of your burdens, and shows
you the way to eternal life. The narrator would have been
library"). Level 3 consisted of explicit interpretations of symbols much better off reading one of his Bibles than the "holy"
or events in the text without mention of interpretive themes (e.g., Book of Sand. But he didn't understand that eternal life comes
"The weight of The Book of Sand symbolizes our guilt in life"). from living, not what we own,
Level 4 was characterized by the interpretation of one or more
theme fragments (i.e., partial themes consistent with the text that I (along with the same trained research assistant who helped score
were not fully elaborated). For example: the written protocols) scored each essay by assigning it to one of
the nine levels of holistic interpretation. There was 92% initial
I think the story is about how we as a society all get wrapped agreement. The remaining 8% of the protocols were discussed and
up in things. Sometimes we put all of our energy on one thing assigned to a level that the research assistant and I agreed on. It is
and forget about others.
worth noting that we never differed by more than one level, with
Level 4 represented an important transition point because readers one exception. In this case, we initially assigned the essay to Levels
went beyond personal opinions, elaborations, and the interpretation 6 and 8, respectively. After discussion, we reassigned the essay to
of isolated symbols to explicitly state a theme. Level 7.
Level 5 consisted of one or more detailed themes (i.e., interpreta-
tion of individual thematic events or episodes without interrelating
them into a coherent holistic interpretation). The following ex- Procedure
ample illustrates an explicit detailed theme:
Individuals participated in groups of 10 to 25 and received
I think the story is saying that we should take things at face identical instructions. I conducted all sessions. There were no time
value and no more. Perhaps the narrator should have looked at limits on any of the experimental tasks.
the book carefully and then left it alone. She was told that one Participants were told they would be reading a text to evaluate
could never find the same illustration twice, but didn't believe
it. She let the book consume her. various aspects of its organization and structure. Each individual
received a packet that included the experimental materials de-
Level 6 represented a basic holistic model, which includes an scribed previously. Individuals were instructed to complete the RBI
explicit theme or integration of one or more main themes without by indicating on a 5-point scale the extent to which they agreed
extensive justification of that theme. For example: with each of the 16 statements. Completion of the RBI required
approximately 3 to 5 min. Next, the participants read a typed,
I believe the meaning deals with priorities. The woman is well double-spaced copy of the story as carefully as possible. Individu-
educated and independent. But once The Book of Sand entered als also were informed that they could reread the text and mark or
her life, everything changed. She became involved with a
mystery she simply could not explain, let alone understand. underline their copy if they chose to. There were no time limits
Part of being is a natural curiosity and desire to comprehend during this phase. The approximate completion time of this stage
what we do not understand. Though this is true, we must not was 5 to 10 min.
allow this to take over our lives. Finding answers to the Next, participants completed the 20-item multiple-choice test
unexplainable is important, but if we lose sight of what is real, without looking back at the story. The average completion time was
we may lose our bearings. approximately 10 min. This test was followed by the written essay
100 SCHRAW

response. Each response booklet contained cover instructions (see Table 1


Appendix C) and two pages of lined paper. The average completion Means and Standard Deviations for Multiple-Choice Tests
time for this phase of the study was approximately 15 to 20 min.
Following this phase, all participants were debriefed. Transaction beliefs
All participants completed trie experimental tasks in the same Low High
order. Materials were not counterbalanced, owing to the logical Transmission
order in which tasks needed to be completed. For example, the RBI beliefs M SD n M SD n
was given prior to reading to avoid confounding due to the Low .76 .12 58 .75 .11 68
experimental text. The multiple-choice test was completed prior to High .74 .14 88 .78 .13 33
the essay to avoid rehearsal and elaboration effects. Because all
participants completed the multiple-choice test under the same
conditions, there is no reason to believe it differentially affected
essay performance. sion scales were used as blocking variables, whereas all
other variables in the study were used as outcome measures.
Scoring
Multiple-Choice Test Performance
The RBI and multiple-choice test were scored objectively. I
(along with the trained research assistant) scored the essays as Table 1 shows the means and standard deviations for the
described previously. multiple-choice recognition test. Neither the main effects for
the transaction and transmission variables nor their interac-
tion reached significance. Participants in all four groups
Results were equally able to read and understand main ideas from
the text, regardless of transmission and transaction beliefs.
Four separate analyses were performed. The first exam-
One possible explanation is that understanding main ideas is
ined the factor structure and internal consistency of the RBI.
necessary, regardless of one's beliefs about the locus of
The second investigated the relationship between reader
meaning. Most college readers presumably possess the skills
beliefs and multiple-choice test performance. The third
and make the effort to understand main ideas in an auto-
examined the relationship between reader beliefs and the 13
mated fashion (McNamara, Kintsch, Songer, & Kintsch,
personal response categories. The fourth considered the
1996).
extent to which reader beliefs are related to holistic text
interpretations. All statistical significance tests are reported
atp < .05. Personal Responses
Several analyses were performed on the 13 response
Factor Analyses of the RBI categories. The first was a 2 X 2 X 3 analysis of variance
(ANOVA), with transaction and transmission scores as
The RBI was analyzed using a principal-axis factor between-subject variables and the mean of responses in the
analysis with a varimax rotation. Two factors emerged that thematic, critical, and personal responses as a three-level
accounted for 71 % of the sample variation in the instrument. within-subject variable. Table 2 shows the means and
The factors, in order of variance explained, were transaction standard deviations associated with this analysis. The main
beliefs (X = 2,068, variance explained = 38%) and transmis- effect for the transaction variable reached significance, F(l,
sion beliefs (\ = 1.906, variance explained = 33%). Indi- 243) = 9.06, MSE = 2.85. Individuals high on the transac-
vidual loadings are given in Appendix A. tion scale generated more responses (M = 3.31, SD — 2.11)
The transaction and transmission scales were orthogonal across the 13 categories than those low on the transaction
(i.e., uncorrelated), suggesting that beliefs on one scale were scale {M = 2.94, SD — 1.83). The repeated measures main
uncorrelated with beliefs on the other scale. This enabled me
to create composite scores for each scale by summing into a
single score the ratings of variables that loaded in excess of Table 2
.40 on that scale. Each scale included six items; thus, each Means and Standard Deviations for Thematic, Critical,
composite score ranged from 0 to 30. The internal consisten- and Personal Responses
cies (Cronbach's alphas) of the transaction and transmission
scales were .81 and ,78, respectively. These findings closely Transaction beliefs
replicate the factor structure and reliabilities reported by Low High
Schraw and Bruning (1996).
beliefs M SD n M SD n
Composite scores for each of the 30-point transaction and
transmission scales were dichotomized on the basis of the Low 58 68
mean of each scale from the present sample and were used to Thematic 3.01 2.58 3.48 2.29
Critical 0.82 0.96 1.01 0.90
create a 2 X 2 grouping matrix in which each individual was Personal 4,87 1.74 5.35 2.12
placed into one of four mutually exclusive categories: high High 88 33
transaction-high transmission, high transaction-low trans- Thematic 2.76 2.02 4.06 4.04
mission, low transaction-high transmission, and low transac- Critical 0.71 0.84 0.94 1.00
Personal 5.44 1.86 5.18 1.60
tion-high transmission. Thus, the transaction and transmis-
READER BELIEFS 101

effect for the three category types (i.e., thematic, critical, and measures ANOVA was performed on the five thematic
personal) also reached significance, F(2, 486) = 251.73, responses. There was a significant main effect for the
MSE = 4.08. A comparison of the three means, using transaction scale, F(\, 243) = 6.35, MSE = 1.32, in which
Fischer's procedure (Levin, Serlin, & Seaman, 1994), re- the high-transaction readers (M = .76, SD = .71) outscored
vealed that more personal responses (M = 5.25, SD = 1.77) the low-transaction readers (Af = .58, SD — .77). There also
were generated than thematic (Af = 3.19, SD = 2.63) or was a significant repeated measures main effect for the five
critical responses (M = .85, SD = .94). Readers spontane- thematic response categories, F(4, 972) = 143.50, MSE =
ously generated significantly more personal responses than .67. A post hoc test using Tukey's Honestly Significant
other categories and generated few critical responses. This Difference Test (HSD) revealed that elaborations (M = 1.87,
may be due to the fact that essay instructions (shown in SD = 1.83) were recalled significantly more than the remain-
Appendix C) encouraged participants to focus on thematic ing four categories. No other effects reached significance.
and personal responses. None of the other main effects or High-transaction readers produced more thematic re-
interactions reached significance. sponses. This was true especially of thematic inferences,
Separate analyses were performed on the thematic, criti- which were generated three times more frequently in the
cal, and personal response categories to examine individual high-transaction group (M = .61, SD = .11) than in the
responses in more detail. Means and standard deviations for low-transaction group (Af=.18, SD = .40). Generating
all 13 categories are shown in Table 3. A 2 X 2 X 5 repeated significantly more themes should be an important indicator
of holistic understanding (Graesser et al., 1994).
An analysis of critical responses showed a significant
Table 3 main effect for the transaction variable, F(l, 243) = 3.23,
Means and Standard Deviations for the MSE - .21, in which the high-transaction readers (M = .33,
13 Response Categories SD = ,56) outscored the low-transaction readers (M = .25,
Transaction beliefs SD = .43). There also was a significant repeated measures
main effect for the three critical response categories, F(2,
Low High 486) = 49.64, MSE = .28. A post hoc test using Tukey's
Transmission
beliefs M SD n M SD n HSD revealed that text-structure responses (M — .59,
SD = .77) were produced significantly more than author-
Low 58 68
purpose ( M = . 1 7 , SD = .35) or text-idea responses
Thematic responses
Retellings 0.44 0.94 0.26 0.70 (M = .09, SD = .27).
Elaborations 1.62 1.37 1.97 1.43 A significant two-way interaction also occurred between
Symbols 0.22 0.49 0.26 0.46 the transaction and transmission variables, F(2,486) = 4.86,
Hypotheses 0.52 0.63 0.44 0.58 MSE — .28. The high-transaction-low-transmission group
Themes 0.20 0.49 0.55 0.72
Critical responses (M = .34, SD = .53) generated significantly more responses
Author's purpose 0.14 0.35 0.13 0.34 than the low-transaction-high-transmission group (Af = .24,
Text ideas 0.12 0.33 0.10 0.30 SD — .47) but did not differ from the high-transaction-high-
Text structure 0.57 0.75 0.78 0.87 transmission (M = .31, SD = .54) or low-transaction-low-
Personal responses transmission groups (Af = .28, SD = .45). High-transaction-
Interest 1.01 0.58 1.18 0.69
Cognitive response 3.63 1.48 3.87 1.56 low-transmission readers were most apt to produce critical
Affective response 0.14 0.57 0.10 0.39 responses, supporting my claim that transaction beliefs are
Events 0.05 0.22 0.05 0.12 related positively to text understanding of narrative text as
Relate to life 0.03 0.18 0.19 0.60 well as findings reported by Wineburg (1991) and Wade et
High 88 33
al. (1994).
Thematic responses
Retellings 0.22 0.63 0.66 1.33 In contrast, an analysis of personal responses failed to
Elaborations 1.96 1.53 1.85 1.91 reveal significant differences, with the exception of a
Symbols 0.13 0.40 0.27 0.51 repeated measures main effect for the five personal re-
Hypotheses 0.30 0.48 0.61 0.70 sponses categories, F(4, 972) = 882.16, MSE = .64.
Themes 0.15 0.39 0.67 0.99
Critical responses Tukey's HSD revealed that cognitive responses (M = 3.83,
Author's purpose 0.16 0.37 0.36 0.70 SD = 3.75) were generated significantly more often than
Text ideas 0.04 0.20 0.12 0.33 interest responses (Af = 1.13, SD — 1.15), which were gen-
Text structure 0.51 0.71 0.45 0.66 erated significantly more often than affective responses
Personal responses
Interest 1.17 0.48 1.18 0.39 (Af = .10, SD = .10), responses relating text to personal life
Cognitive response 3.98 1.80 3.76 1.28 (M = .11, SD = .11), or responses relating to events de-
Affective response 0.06 0.23 0.12 0.33 scribed in the text (M = .06, SD = .07).
Events 0.13 0.40 0.08 0.17 The main effects for the transaction and transmission
Relate to life 0.11 0.32 0.09 0.29
variables did not reach significance in the analysis of
Note. Scores in each category, summed across participants, range personal responses. This finding was unexpected given the
as follows: retellings, 0-4; elaborations, 0-7; symbols, 0-3; assumption that transactional reading should increase per-
hypotheses, 0-2; themes, 0-5; author's purpose, 0-2; text ideas,
0-3; text structure, 0-3; interest, 0-5; cognitive responses, 0-8; sonal engagement and empathy. One possibility is that The
affectiveresponses,0-3; events, 0-2; and relate to life, 0-2. Book of Sand was interesting enough to engage most
102 SCHRAW

readers, regardless of the strength of transaction and transmis- Table 5


sion beliefs. A second possibility is that some participants Response Frequencies Among the Nine Levels
felt uncomfortable or unable to respond to a highly interpre- of Holistic Interpretation
tative text on relatively short notice. An alternative explana-
Level of holistic interpretation
tion is that transaction beliefs have no impact on personal Transaction
responses, a view that is strongly at odds with current reader belief 0 1 2 3 4 5 6 7 8
response theory (Rosenblatt, 1994; Squire, 1994; Tompkins, Low 15 64 30 25 4 8 0 0

OO
1980). Further research is needed to replicate this effect and High 14 22 21 11 15 14 3 1
to determine whether transaction beliefs are less important
than thought.

Holistic Text Interpretations blatt, 1994; Tompkins, 1980) that transaction beliefs pro-
mote higher level meaning construction, particularly when
Overall understanding of the text was assessed using a reading a highly interpretable narrative text. This assertion
9-level holistic interpretation score. Means and standard has not been tested directly. A second reason is to gain a
deviations are shown in Table 4. A 2 X 2 ANOVA, with the better understanding of the interpretative strategies (e.g.,
transmission and transaction scales as between-subject inde- thematic inferences) readers use when constructing meaning
pendent variables, revealed a significant main effect for the from a narrative text.
transaction scale, F(l, 243) = 5.99, MSE = 1.95. High-
I distinguished between transmission and transaction
transaction readers generated more sophisticated holistic
beliefs. The transmission model views reading as a process
interpretations {M — 3.00, SD = 1.64) compared with low-
of transmitting meaning from the author to the reader's
transaction readers (Af = 2.51, SD = 1.18). No other effects
memory. The transaction model views reading as a process
reached significance.
that emphasizes a reciprocal transaction among the reader,
The frequency of each of the nine levels of holistic text, and author. On this view, readers construct meaning on
interpretation for high- and low-transaction readers are the basis of information in the text, perceived authorial
presented in Table 5. Low-transaction readers consistently intent, and, most important, their own individual goals and
scored at Levels 0 through 3, whereas high-transaction knowledge.
readers were more frequently in the Level 4 to Level 7 range.
My general hypothesis was that readers characterized by
None of the readers scored at Level 8. Summarizing the data
strong transaction beliefs construct more sophisticated mean-
in Table 5, low-transaction readers were twice as likely as
ings than those characterized by strong transmission beliefs.
high-transaction readers to generate low-level holistic inter-
I used this hypothesis to address the following four research
pretations, whereas high-transaction readers were three
questions:
times more likely to generate higher level interpretations.
Regardless of their beliefs, however, few readers produced 1. Is there a statistical relationship between the transmis-
sophisticated holistic interpretations. It is not clear whether sion and transaction scales included in the RBI?
this was due to the inherent difficulty of interpreting a text or 2. Are transmission and transaction beliefs related to
to limitations imposed by the experiment itself. multiple-choice test performance?
3. Are transmission and transaction beliefs related to
individual interpretative responses?
Discussion 4. Are transmission and transaction beliefs related to
The purpose of this research was to explore the relation- holistic text interpretations?
ship between adults' beliefs about the meaning construction The first question examined the relationship between the
process and the construction of text meaning when reading a two hypothesized scales. An exploratory factor analysis
highly interpretable narrative text. This question is impor- reported two uncorrelated (i.e., orthogonal) factors that were
tant for two reasons. One is to test the claim endorsed by internally consistent and explained a meaningful proportion
reading response theorists (Farrell & Squire, 1990; Rosen- of sample variation. This finding replicates that of Schraw
and Bruning (1996) who reported two uncorrelated factors
similar to those observed here and is consistent with claims
made by reader response theorists (Bogdan & Straw, 1990;
Table 4 Tompkins, 1980). I concluded that the transmission and
Means and Standard Deviations for transaction scales constitute separate beliefs about the
Holistic Interpretation Scores reading process; that is, readers hold beliefs about the
TYansaction beliefs importance of the author's intended meaning that are
Low High
independent of their beliefs about the reader's role in
Transmission meaning construction. Each of these beliefs may affect
beliefs M SD n M SD n reading outcomes differently.
Low 58 68 My second research question examined whether recogni-
2.65 1.43 3.02 1.67 tion test performance for main ideas differed among these
High 88 33 four groups. Schraw and Bruning (1996) reported signifi-
2.40 1.01 2.98 1.63
cantly higher recall for both text propositions and proposi-
READER BELIEFS 103

tional modifiers among high-transaction readers. However, higher were high-transaction readers. Unfortunately, fewer
there were no recognition test differences in the present than 20% of all readers scored at Level 4 or above, and only
study. One explanation is that the present sample included a 4 readers (i.e., less than 2%) scored at Level 6 or above.
large number of skilled college readers who are adept at Most readers failed to generate a coherent thematic sum-
identifying main ideas, perhaps in an automated fashion. mary of the story at all (i.e., scored lower than Level 4). One
Beliefs in general, and transaction beliefs in particular, may possible reason is that interpreting an abstract, symbolic
not affect automated comprehension skills such as under- narrative story is a difficult task. A second reason is that
standing main ideas (Graesser et al., 1994; Kintsch, 1998). readers were given approximately 20 rain to do so, preclud-
McNamara et al. (1996), for example, found that the benefits ing the opportunity to discuss and reflect on the story's
of active processing are gained primarily at the global level possible meanings. Overall, the present results indicate that
of comprehension (e.g., holistic interpretations) rather than it is those with strong transaction beliefs who primarily are
at surface (e.g., phonological access) or text-based (e.g., willing or able to construct a sophisticated holistic interpre-
main ideas) levels. tation of a text.
My third question addressed whether transmission and These findings lead to two main conclusions. The first is
transaction beliefs are related to individual interpretative that adult readers hold separate beliefs (i.e., transmission
responses. I predicted that transaction beliefs would increase and transaction) that affect their reading in different ways.
the number of thematic, critical, and personal responses that Transaction beliefs were related to several measures of
were generated in readers' essays. Thematic and critical constructed meaning, namely, thematic responses, critical
responses increased significantly, whereas personal re- responses, and holistic interpretations. In contrast, transmis-
sponses did not. High-transaction readers generated three sion beliefs were unrelated to all outcome measures in this
times more thematic inferences than low-transaction read- study, with one exception: Transmission and transaction
ers. In the critical response category, an interaction was beliefs interacted in the analysis of critical responses. In this
observed between the transmission and transaction variables case, transmission beliefs had a negative impact on re-
in the expected direction. High-transaction-low-transmis- sponses.
sion readers generated more critical responses than did A second conclusion is that transaction beliefs are related
low-transaction-high-transmission readers. These findings to deeper processing in the form of thematic and critical
are of special importance because they reveal that transac- responses and to the construction of higher level holistic
tion beliefs may enhance the meaning construction process interpretations in narrative text. Deeper constructive process-
in narrative text by increasing the type and amount of ing is consistent with reader response theories (Rosenblatt,
thematic and critical engagement. Traditionally, reader re- 1994; Squire, 1994; Tompkins, 1980) as well as psychologi-
sponse theorists have argued that transaction beliefs increase cal text-processing theories (Graesser et al., 1994; Kintsch,
personal engagement and responses, especially when relat- 1998). In contrast, neither transmission nor transaction
ing text information to one's personal life and empathizing beliefs were related to recognition test performance, which
with characters. Surprisingly, these effects were not ob- is thought to measure text-based comprehension rather than
served in the present study. One explanation is that The Book higher level construction of the text's holistic meaning
of Sand invoked spontaneous interest and engagement in a (McNamara et al., 1996).
majority of readers, thereby minimizing the effects of strong
The present findings support the idea that transaction
transaction beliefs. The data in Table 3, however, do not
beliefs are related to the meaning construction process when
strongly support this view, as few readers generated affective
reading a highly interpretable narrative text. This claim is
responses of related text information to their personal lives.
consistent with response theory (Rosenblatt, 1994; Tomp-
An alternative explanation is that The Book of Sand includes
kins, 1980) and previous empirical findings that demon-
content and themes that were difficult for readers to relate to
strated a relationship between transaction beliefs and recall
personally. One possibility is that the dark overtones of the
of text information (Schraw & Bruning, 1996). Transmis-
story discouraged personal involvement without affecting
sion beliefs, even when they are strong, have little effect on
critical responses to the text or thematic responses geared
reading outcomes. Although four distinct configurations
toward understanding. Future research is needed in which
were possible in the present study (i.e., high or low on the
The Book of Sand is compared with another text that is easier
transmission and transaction scales), the important distinc-
for readers to engage in personally. Such a study would help
tion appears to be whether a reader endorses strong transac-
determine whether the relative lack of personal engagement
tion beliefs. Those with strong transaction beliefs generate
in the present study was due to the story itself or to reader
more responses and construct more sophisticated holistic
characteristics.
interpretations, regardless of their transmission beliefs.
My fourth question addressed whether transmission and The present results shed new light on the relationship
transaction beliefs are related to holistic text interpretations. between reader beliefs and understanding narrative text.
High-transaction readers constructed higher level holistic Transaction beliefs appear to facilitate higher level meaning
interpretations than did low-transaction readers. This finding construction more than transmission beliefs. However, these
is consistent with reader response theories and suggests that findings may not apply to the meaning construction process
transaction beliefs may have their greatest influence on when reading expository texts. Future research should
higher level meaning construction aspects of reading. Table compare meaning construction strategies across a variety of
5 reveals that the majority of readers scoring at Level 4 or texts that differ by genre, content, and difficulty.
104 SCHRAW

The present findings should be tempered as well, given Many, J., & Cox, C. (1992). Reader stance and literary understand-
that most of the analyses reported relatively small effect ing. Norwood, NJ: Ablex.
sizes, usually .30 or below. Effect sizes in this range should Many, J. E., & Wiseman, D. L. (1992). The effect of teaching
be interpreted with caution because they suggest that other approach on third-grade students' response to literature. Journal
of Reading Behavior. 24, 265-288.
variables may explain significant proportions of variance in
Mathewson, G. C. (1994). Model of attitude influence upon reading
outcome measures (Cohen, 1992). It appears that transaction and learning to read. In R. B. Ruddell, M. R. Ruddell, & H.
beliefs make a significant yet modest contribution to the Singer (Eds.), Theoretical models and processes of reading (4th
meaning construction process. Although not examined here, ed., pp. 1131-1161). Newark, DE: International Reading Asso-
it is possible that transaction beliefs also interact with other ciation.
beliefs (e.g., self-efficacy) or prior knowledge. McNamara, D. S., Kintsch, E., Songer, N. BM & Kintsch, W.
(1996). Are good texts always better? Interactions of text
coherence, background knowledge, and levels of understanding
References in learning from text. Cognition and Instruction, 14, 1-43.
Rosenblatt, L. M. (1994). The transactional theory of reading and
Alexander, P. A., & Dochy, F. J. R. C. (1994). Adults' views about writing. In R. B. Ruddell, M. R. Ruddell, & H. Singer (Eds.),
knowing and believing. In R. Garner & P. A. Alexander (Eds.), Theoretical models and processes of reading (4th ed., pp.
Beliefs about text and instruction with text (pp. 223-244). 1057-1091). Newark, DE: International Reading Association.
Hillsdale, NJ: Erlbaum. Schraw, G. (1997). Situational interest in literary text. Contempo-
Bogdan, D. (1990). In and out of love with literature. In D. Bogdan rary Educational Psychology, 22, 436-456.
& S. Straw (Eds.), Beyond communication: Reading comprehen- Schraw, G., & Bruning, R. (1996). Readers' implicit models of
sion and criticism (pp. 109-138). Portsmouth, NH: Heinemann. reading. Reading Research Quarterly, 31, 290-305.
Bogdan, D., & Straw, S. (Eds.) (1990). Beyond communication; Schunk, D. H., & Zimmerman, B. J. (1997). Developing self-
Reading comprehension and criticism. Portsmouth, NH: Heine- effkacious readers and writers: The role of social and self-
mann. regulatory processes. In J. T. Guthrie & A. Wigfield (Eds.),
Borges, J. L. (1977). The book of sand. (N. T. Di Giovanni, Trans.). Reading engagement: Motivating readers through integrated
New York: E. P. Dutton. (Original work published 1969) instruction (pp. 34-50). Newark, DE: International Reading
Chambliss, M. (1994). Why do readers fail to change their beliefs Association.
after reading persuasive text? In R. Garner & P. A. Alexander Shell, D. F, Murphy, C , & Bruning, R. H. (1995). Self-efficacy,
(Eds.), Beliefs about text and instruction with text (pp. 75-89). attribution, and outcome expectancy mechanisms in reading and
Hillsdale, NJ: Erlbaum. writing achievement: Grade-level and achievement-level differ-
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, ences. Journal of Educational Psychology, 87, 386-398.
155-159. Squire, J. R. (1994). Research in reader response, naturally
Dole, J. A., & Sinatra, G. A. (1994). Social psychology research on interdisciplinary. In R. B. Ruddell, M. R. Ruddell, & H. Singer
beliefs and attitudes: Implications for research on learning from (Eds.), Theoretical models and processes of reading (4th ed., pp.
text. In R. Garner & P. A. Alexander (Eds.), Beliefs about text 637-652). Newark, DE: International Reading Association.
and instruction with text (pp. 245-265). Hillsdale, NJ: Erlbaum. Straw, S. (1990). Challenging communication. In D. Bogdan & S.
FarreU, E. J., & Squire, J. R. (Eds.). (1990). Transactions with Straw (Eds.), Beyond communication: Reading comprehension
literature. Urbana, IL: National Council of Teachers of English. and criticism (pp. 67-90). Portsmouth, NH: Heinemann.
Garner, R., & Alexander, P. A. (1994). Beliefs about text and Straw, S., & Sadowy, P. (1990). Dynamics of communication. In D.
instruction with text. Hillsdale, NJ: Erlbaum. Bogdan & S. Straw (Eds.), Beyond communication: Reading
Garner, R., & Hansis, R. (1994). Literacy practices outside of comprehension and criticism (pp. 21-48). Portsmouth, NH:
school: Adults' beliefs and their response to "street texts." In R. Heinemann.
Gamer & P. A. Alexander (Eds.), Beliefs about text and Thomson, J. (1993). Helping students control texts: Contemporary
instruction with text (pp. 57-74). Hillsdale, NJ: Erlbaum. literary theory into classroom practice. In S. Straw & D. Bogdan
Graesser, A. C , Singer, M., & Trabasso, T. (1994). Constructing (Eds.), Constructive reading: Teaching beyond communication
inferences during narrative text comprehension. Psychological (pp. 130-154). Portsmouth, NH: Heinemann.
Review, 101, 1-25. Tompkins, J. P. (Ed.). (1980). Reader-response criticism: From
Harste, J. C , Burke, C. L., & Woodward, V. A. (1994). Children's formalism to post-structuralism. Baltimore, MD: Johns Hopkins
language and world: Initial encounters with print. In R. B. Press.
Ruddell, M. R. Ruddell, & H. Singer (Eds.), Theoretical models Wade, S. E., Thompson, A., & Watkins, W. (1994). The role of
and processes of reading (4th ed., pp. 48-69). Newark, DE: belief systems in authors' and readers' constructions of text. In R.
International Reading Association. Garner & P. A. Alexander (Eds.), Beliefs about text and
Hartman, D. K. (1995). Eight readers reading: The intertextual instruction with text (pp. 265-294). Hillsdale, NJ: Erlbaum.
links of proficient readers reading multiple passages. Reading Weber, R. P. (1985). Basic content analysis. Newbury Park, CA:
Research Quarterly, 30, 520-561. Sage.
Hunt, R. A. (1990). The parallel socialization of reading research Wineburg, S. S. (1991). On the reading of historical texts: Notes on
and literary theory. In D. Bogdan & S. Straw (Eds.), Beyond the breach between school and academy. American Educational
communication: Reading comprehension and criticism (pp. Research Journal, 28, 495-520.
91-105). Portsmouth, NH: Heinemann. Zwann, R. A. (1994). Effect of genre expectations on text
Kintsch, W. (1998). Comprehension: A paradigm for cognition. comprehension. Journal of Experimental Psychology: Learning,
Cambridge, England: Cambridge University Press. Memory, and Cognition, 20, 920-933.
Levin, J. R., Serlin, R. C , & Seaman, M. A. (1994). A controlled, Zwann, R. A., & Radvansky, G. A. (1998). Situation models in
powerful multiple-comparison strategy for several situations. language comprehension and memory. Psychological Bulletin,
Psychological Bulletin, 115, 153-179. 123, 162-185.
READER BELIEFS 105

Appendix A

Reader Belief Inventory

1. Good readers remember most of what they read verbatim (DNL) 10. I enjoy interpreting what I read in a personal way
2. I like the fact that two people can read the same book and (TA; .64)
disagree about what it means (TA; .42) 11. Reading for pleasure is the best kind of reading (TA; .46)
3. The main purpose of reading is to understand what the author 12.1 like books where you know exactly what the author means
says (TM; .52) (TM; .50)
4.1 like to make judgments about the author's writing style when 13.1 enjoy sharing the thoughts and reactions of characters in a
I read (DNL) book with others (TA; .55)
5. When I read, I try to carry away exactly what the author meant 14. When I read, I focus on what the author says is important
(TM;.6O) (TM;56)
6.1 often have strong emotional responses to what I read (TA; .52) 15. Most books mean exactly what they say (TM; .51)
7. I like poetry more than technical text because it is more 16. When I read, I focus more on how I feel about the
expressive (DNL) information than on what I learn (DNL)
8. People should agree on what a book means (TM; .45) Note. Numbers in parentheses indicate factor loadings. DNL -
9. When I read, I like to imagine I am living through the item did not load in excess of .40 on a factor; TA = transaction
experience too (TA; .58) beliefs; TM = transmission beliefs.

Appendix B

Sample Test Questions


The salesman introduced The Book of Sand as a The salesman indicated
a) encyclopedia a) the book was printed before the 19th century
b) holy book* b) the book was printed in the 19th century
c) "curiosity" c) the book was printed in the 20th century
d) sacred text d) he did not know when the book was printed*
Note. Asterisks indicate correct responses.

Appendix C

Instructions for Essay

Each time a person reads a story, he or she has some kind of means and (b) what kind of thoughts and emotions it evoked in you.
personal reaction to it. In this part, we want you to describe your Don't worry about whether your response is correct or not; there are
personal response to the story The Book of Sand. First we want you norightor wrong answers. Please be informal; we are most interested in
to reread the story. Next we want you to write a two-page description of what you experienced when you read the story. Everything you write is
your response. Please try to describe (a) what you think the story completely anonymous, so don't worry about being candid.

{Appendixes continue)
106 SCHRAW

Appendix D

Thirteen Response Subcategories

Thematic Responses 8. Critical analysis of text structure: Statements pertaining to the


organization and coherence of the text
1. Retellings: Reiterating information included in the text
2. Text-based elaborations: Elaboration of information stated in
text
3. Interpretation of symbols: Assigning a symbolic meaning to a Personal Responses
character or event in the text
9. Engagement and interest: Statements pertaining to engage-
4. Generate hypotheses and explanations: Providing a coherent
ment, interest, liking, or curiosity about the text
rationale for information and themes mentioned in text
10. Cognitive reactions: Statements pertaining to the reader's
5. Thematic statements: Statements pertaining to major themes
thoughts and intellectual reactions to the text
included the text
11. Affective reactions: Statements pertaining to emotional
reactions or feelings created by the text
Critical Responses 12. Relate to events or characters: Statements in which the reader
6. Author's purpose and intention: Statements about the author's attempts to understand or empathize with events or characters in
purpose or presumed intentions for writing the text the text
7. Critical analysis of text ideas: Statements critiquing informa- 13. Relate to personal experiences: Statements in which the
tion included in the text, but especially its clarity, descriptiveness, reader relates information in the text to his or her own personal
and overall effectiveness experiences

Appendix E

Nine Levels of Holistic Interpretation

0 = Unwilling or unable to interpret text at any level integration of one or more main themes without explicit justifica-
1 = Personal reactions and opinions without any text-based tion of that theme)
elaborations, interpretation of symbols, or thematic inferences 7 = Elaborated model (i.e., comprehensive interpretation or
2 = Text retelling and elaborations only, without clarification of comparison of one or more interpretations; includes some explicit
symbols or themes justification of themes)
3 = Interpretation of symbols, events, or characters only, but no 8 = Integrated model (i.e., relates elaborated interpretative model
thematic interpretation to personal experiences or philosophy, or to another text)
4 = Interpretation of one or more theme fragments (i.e., partial
themes consistent with the text)
5 = Interpretation of one or more detailed themes (i.e., interpre-
tation of individual thematic events or episodes without interrelat- Received November 3,1998
ing them into a coherent thematic model) Revision received May 27,1999
6 = Basic model (i.e., clear presentation of a main theme or Accepted May 27, 1999

You might also like