A Review Of Reading Motivation Scales

5m ago
4 Views
0 Downloads
607.31 KB
80 Pages
Last View : 8d ago
Last Download : n/a
Upload by : Luis Wallis
Transcription

A Review of Reading Motivation ScalesMarcia H. DavisJohns Hopkins [email protected] M. TonksNorthern Illinois [email protected] HockUniversity of KansasWenhao WangUniversity of KansasAldo RodriguezNorthern Illinois UniversityTo appear in Reading Psychology:Davis, M. H., Tonks, S. M., Hock, M., Wang, W., & Rodriguez, A. (2018). A review of readingmotivation scales. Reading Psychology, 39, 82Reading Psychology peer review process: Manuscripts submitted to this journal undergo editorialscreening and peer review by anonymous reviewers.Corresponding Author:Marcia H. Davis ([email protected])Center for Social Organization of Schools2701 North Charles Street, Suite 300Johns Hopkins UniversityBaltimore, MD 21218Authors’ NoteThe present research was supported by grants awarded to the authors by the Institute ofEducation Sciences, U.S. Department of Education (R305A110148 and R305A150193). Allopinions expressed in this manuscript are those of the authors and not the funding agencies.

2AbstractReading motivation is a critical contributor to reading achievement and has the potential toinfluence its development. Educators, researchers, and evaluators need to select the best readingmotivation scales for their research and classroom. The goals of this review were to identify a setof reading motivation student self-report scales used in research, examine the development andpsychometric properties of each reading motivation scale, and compare scales on availability,reliability, age range, and motivation constructs measured. This article summarizes 16 currentreading motivation scales. Findings suggest the need for more research regarding themultidimensionality of reading motivation and measures that could span early childhood toadolescence.

3A Review of Reading Motivation ScalesWhile proficiency in reading is critical to understanding core class texts, researchers alsobelieve that reading skills and strategies do not fully account for the variability in students’engagement in reading. In fact, students engage or disengage in reading for a variety of reasons.For example, they may enjoy the process of reading or believe that reading is a valuable way tolearn information. Students who disengage from reading, however, may not lack the ability toread but resist reading due to a lack of motivation. This disengagement will eventually havedetrimental effects on their reading ability (Baker, Afflerbach, & Reinking, 1996; Guthrie &Wigfield, 1999; Guthrie, McGough, Bennett, & Rice, 1996; Paris & Oka, 1986). In addition,reading motivation continually surfaces as a critical contributor to reading achievement(Biancarosa & Snow, 2004; Curtis, 2002). Reading motivation is highly related to readingcomprehension and achievement (Anmarkrud & Braten, 2009; Cartwright, Marshall, & Wray,2016; Law, 2009; Mucherah & Yoder, 2008; Park, 2011; Retelsdorf, Koller, & Möller, 2011;Wang & Guthrie, 2004) and has also been shown to predict later reading achievement (Becker,McElvany, & Kortenbruck, 2010; Schaffner, Philipp, & Schiefele, 2016; Taboada, Tonks,Wigfield, & Guthrie, 2009). Thus, reading skills and reading motivation seem inextricably linked(e.g., Adelman & Taylor, 2000; Ellis et al., 1997; Zins et al., 2004). Studies of this relationshipmay be key to improving reading achievement, particularly for students who struggle withreading proficiency. Fundamental to these studies is the ability to develop and identify valid andreliable scales of reading motivation. The overarching purpose of this article is to describe andreview current reading motivation measures.While there are multiple definitions of reading motivation, Guthrie and Wigfield’s (2000)definition of reading motivation as an “individual’s personal goals, values, and beliefs with

4regard to the topics, processes, and outcomes of reading” (p. 405) seems to best personalizereading motivation. According to this definition, reading motivation differs at an individuallevel. An individual’s reading motivation may also differ depending on context, such as schooland home (De Naeghel et al., 2012; McKenna & Kear, 1990) and by the way the text ispresented, such as print or digital (McKenna et al., 2012). Most will also agree that readingmotivation is multifaceted and complex and may include sub-constructs such as intrinsic readingmotivation, extrinsic reading motivation, social reasons for reading, and value of reading(Wigfield & Guthrie, 1997).A recent review written by leading researchers in reading motivation detailed howconstructs of reading motivation have been defined, clarified the dimensionality of theconstructs, and reviewed research on the relationships among motivation constructs and otherreading variables such as reading behavior and competence (Schiefele, Schaffner, Möller, &Wigfield, 2012). It is the author’s belief that two separate groups of reading motivationconstructs exist. The first group, relating to intrinsic and extrinsic reasons for reading, includesreading attitude, intrinsic value, and reading value. Therefore, a reader who is intrinsicallymotivated to read may read out of enjoyment of reading, the value of reading, or a positiveattitude towards reading. A reader who is extrinsically motivated to read may be motivated byexternal sources such as grades or recognition. The second group of reading motivationconstructs includes self-concept of reading ability and reading self-efficacy. These constructs areactually antecedents of reading motivation since they describe the “expectancy of successfulreading” (Schiefele et al., 2012, p. 431).Because reading motivation is a critical contributor to reading achievement and has thepotential to influence its development, researchers have developed multiple scales to measure

5reading motivation. Although there are various forms of reading motivation scales, such asparent beliefs about their children’s reading motivation (Wigfield & Guthrie, 1997) and teacherreports (Wigfield et al., 2008), the most commonly used assessment of reading motivation isstudent self-report. Unlike other forms of measurement, self-reports are fairly easy and quick toassess in a classroom and rely less on adult time, but they are not without their difficulties andbiases (Brenner & DeLamater, 2016). Because of their prevalence, it is important to documentwhat self-report scales are available for measuring reading motivation as well as the reliabilityand validity of these scales.Goals of Current ReviewReading motivation scales vary as to the reading constructs they measure and theappropriateness of the measure for different aged readers. However, the last review of readingscales related to affective responses to reading was by Summers (1977). Summers stated that“cognitive characteristics may determine the limits to a student’s development but affectivecharacteristics will influence whether or not the attempt is made to reach those limits”(Summers, 1973, p.2). In his review, Summers summarized the item development, response setcontamination, and the statistical procedures including reliability and information on factoranalyses of scales measuring affective responses to reading, reading attitude scales, publishedmostly in the late 1960s. At that time, only a few standardized tools existed for measuringaffective states related to reading such as reading attitude. Summers noted after reviewing thesescales that most had problems including restricted sampling, response set bias, item analysis, andinformation about reliability and validity.In the 1970s, educators, researchers, and evaluators expressed a need to have appropriatereading motivation measures for their goals. However, although the reviews of reading

6motivation measures discussed the theoretical constructs of reading motivation (Schiefele et al.,2012) and terminology (Conradi, Jang, & McKenna, 2014), there has not been a systematicreview of reading motivation measures since 1977. The goal of the current review was to identifyand systematically evaluate common and recently developed reading motivation scales. It wasour belief that an up-to-date review of reading motivation scales would not only help individualsselect appropriate measures but would also highlight appropriate next steps for readingmotivation scale development.MethodLiterature Search and Review CriteriaThe objective of the search for reading motivation scales was to identify as many selfreport reading motivation scales found in the literature for all age levels. The databases searchincluded PsychInfo, Psych Articles, and Eric using keywords related to reading motivation andmeasurement (i.e., reading, motivation, comprehension, self-efficacy, self-concept, attitude,value, confidence, measurement, and scale). In addition to online searches, the team examinedreference sections of chapters in books on the topic of reading motivation and other identifiedstudies as well as conference proceedings of the American Educational Research Association.Further, the team completed an additional search using the name of each scale as a keyword inorder to examine the extent to which each measure was used in past research.Reading motivation scales had to meet several requirements to be included in this review.First, the motivation scales were limited to scales after and including the year 1990. Second, welimited scales to those that included some aspect of a self-report Likert-type scale. The scaleswere not limited by respondent age and included a review of scales appropriate for respondentsin first grade to adult. Over 200 papers including a reading motivation assessment were

7identified, however only a total of 120 articles were identified that included the use of a selfreport reading motivation measures with a Likert-type scale developed on or after 1990. Byorganizing the 120 papers into stacks of those using the same scale or adaption of the same scale,a total of 16 student self-report scales of reading motivation and self-efficacy were found. A fewpopular scales, or adaptions of these popular scales, were used in a majority of the papers. Thelist of sixteen scales included: The Children’s motivations for Reading Scale, Young ReaderMotivation Questionnaire, Young Children’s Academic Intrinsic Motivation Inventory, TheChildren’s Academic Intrinsic Motivation Inventory, Elementary Reading Attitude Survey,Motivation to Read Profile, Motivation for Reading Questionnaire, Reader Self-PerceptionScale, Reading Self-Concept Scale, the SRQ-Reading Motivation Questionnaire, The AdaptiveReading Motivation Measure, Adolescent Motivation to Read Profile, Adult Motivation forReading Scale, Motivations for Reading Information Books, Reader Self-Perception Scale 2, andThe Survey of Adolescent Reading Attitudes.Each scale was reviewed separately on the following criteria as outlined in Tables 1-3:accessibility of the full scale, the age range of the developmental sample, languages available,number and label of constructs measured, reliability of full scale and subscales, size ofdevelopmental sample, and information on construct validity. These particular characteristicswere reviewed and included in the three tables since they provide the information needed inorder for an individual to determine scale appropriateness. Reliability was assessed usingmagnitudes recommended by Sattler (2006) where coefficients above .80 are considered reliable,.70-.79 relatively reliable, .60-.69 marginally reliable, and below .59 unreliable. When the datawere available, race, gender, and grade differences on the measures were included in theindividual descriptions in order to indicate that the scale was measuring reading motivation in a

8similar way for different groups, which is a type of construct validity. Two researchers and aresearch assistant read each development paper and wrote descriptions and critiques of eachscale. This information is summarized in Tables 1, 2, and 3. Inter-rater reliability was not neededsince the information was pulled directly from the development papers.For each early childhood, elementary, and adolescents/adults scale, the current reviewpresents an overview, the constructs measured with reliability information when available, thedevelopment and validation of the scale, and a critique of the scale. A discussion following theindividual descriptions describes general findings across all sixteen scales. Table 1 about here Table 2 about here Table 3 about here ResultsReading Motivation Scales: Early ChildhoodThree of the reading motivation scales fell into an early childhood category. Theseinclude the Children’s motivations for Reading Scale (CMRS), Young Reader MotivationQuestionnaire (YRMQ), and Young Children’s Academic Intrinsic Motivation Inventory (YCAIMI). A fourth measure that could be used for early childhood, Elementary Reading AttitudeSurvey (ERAS), is described in the elementary section since it can be used with students fromfirst to sixth grades. In this section, a short description and review of the three early childhoodreading motivation scales is included.Children’s Motivations for Reading Scale. CMRS (Baker & Scher, 2002) was writtento assess the multidimensionality of reading for beginning readers. The developers wanted to gobeyond other scales that focused primarily on enjoyment of reading to tap into motivational

9constructs of enjoyment, value, and perceived competence that had been found in research. Theitems were selected from other reading motivation scales such as the Heathington Primary Scale(Alexander & Filler, 1976), the Survey of Reading Attitudes (Alexander & Engin, 1986), theEstes Attitude Scale (Estes, 1971), and a series of inventories developed by Gambrell et al.(1996) (Baker & Scher, 2002, p. 246). According to Baker and Scher (2002) the scale waswritten in order to measure motivation for reading in beginning readers. The scale was developedusing responses from first grade students, and therefore, would be appropriate to use in a firstgrade classroom. The full scale and a description of its development can be found in Baker andScher (2002). The information in the following two sections was obtained from this developmentpaper.Constructs Measured. The 16 item CMRS (α .86) includes 7 items measuringenjoyment in reading (I like to read, α .67), 4 items measuring value of reading (I think peoplecan learn new things from books, α .78), 3 items measuring perceived competence in reading (Ithink I will be a good reader, α .67), and 2 items measuring and library-related topics (I like togo to the school library, α .57). Students respond on a 4-point scale. To administer the survey,an adult shows the student two stuffed animals: One animal agrees and one does not agree with astatement such as “Regal thinks books are good places to find answers to questions but Cha Chadoesn’t think books are good places to find answers to questions. Who are you more like?” (p.246). Next, the student is asked if they are “a lot” or “just a little” like the stuffed animal.Development and Validation. Baker and Scher administered the survey to 65 first gradestudents from six school in Baltimore. A principal components analysis with Varimax rotationwas conducted on the original twenty items which indicated five factors, although only threefactors were interpretable (value, enjoyment, and perceived competence). They caution

10interpretation of these findings based on limited power due to a small sample size. They alsofound in the analyses that children rated the enjoyment scale significantly lower than the value orperceived competence scales.In addition to children taking CMRS, parents were asked about their children’sexperiences with reading at home and about their own beliefs about the importance of reading.When examining correlations among the parent ratings and student scores, they found positiveand statistically significant correlations between the parental reports of students’ interest inlearning to read and parental view that reading is a course of pleasure with the total studentreported motivation as well as with the subscales of enjoyment, value, and perceivedcompetence.Critique. Overall, CMRS is a somewhat reliable and valid measure of reading motivationfor first grade students. Strengths of CMRS include that it has developmentally appropriatequestions and that has been shown to relate well to other reading motivation behaviors such asinterest in reading. In addition, with only 16 items it should not take long to administer to anindividual child. Despite these strengths there are several weaknesses of CMRS. The scale isindividually administered, and therefore, takes considerable time to administer to a whole class.CMRS was only validated with a very small sample of first grade students and has not been usedin other published papers; therefore, it is difficult to determine how well the scale would performwith students in other grades or demographics. Finally, some constructs are composed of onlytwo to three items, which may not capture all aspects of the constructs and a few constructsreliabilities were lower than .70.Young Children’s Academic Intrinsic Motivation Inventory. To study the academicintrinsic motivation of younger students, Gottfried (1990) introduced Y-CAIMI. Gottfried

11believed that it was important to study motivation in younger elementary school students since,at the time, there was little to know information on young children’s academic intrinsicmotivation. Second, Gottfried believed that motivation in early elementary years could havelasting implications for future success in school. To develop Y-CAIMI, Gottfried modifiedCAIMI (discussed below) developed for older elementary school students by reducing thenumber of items and writing a simpler response format. The scale was validated for use withstudents between seven and nine years of age. Sample items and a description of its developmentcan be found in Gottfried (1990). The information in the following two sections was obtainedfrom this development paper.Constructs Measured. The 39 item Y-CAIMI (α .91) includes 12 items measuringintrinsic motivation for reading (I like learning new things in reading, α .82), 12 itemsmeasuring intrinsic motivation for math (I like learning new things in math, α .84), 12 itemsmeasuring intrinsic motivation for general learning (I like learning new work in school, α .82),and 3 items measuring preference for difficult work (I like to do easy work, α .87). Studentsrespond on a three-point Likert scale including responses of Very True, A Little True, and NotTrue. Y-CAIMI is individually administered. An adult reads all items and response choicesaloud and the student points to an index card with his or her response choice. Administration ofY-CAIMI takes 20-30 minutes.Development and Validation. Gottfried administered Y-CAIMI to 107 children over acourse of three years, covering the time span of when the children were 7 to the time they were 8(roughly grades 1-2 to grade 2-3). At age 9 they were administered CAIMI. Confirmatory factoranalysis was used to examine the factor structure, indicating that the best fit was a four-factormodel (reading, math, general motivation, and enjoyment of difficult schoolwork). In addition to

12Y-CAIMI, students were administered assessments of IQ and achievement. In relation to thesubscale, intrinsic motivation for reading, there were no significant differences among grades orgender. Y-CAIMI reading scores at age 8; however, correlated positively and significantly toCAIMI reading scores at age 9. At age 7 there was a significant and positive relationshipbetween both IQ and reading achievement with Y-CAIMI reading score.Critique. Overall, Y-CAIMI is a reliable and valid scale of intrinsic reading motivationfor first, second, and third grade students. Strengths of Y-CAIMI include having been adaptedfrom an earlier established measure for older students, confirmatory factor analysis showing astable factor structure, reliability alpha scores greater than .80 for all subscales, Y-CAIMI hasbeen translated into Spanish (Touron, Reparaz, & Peralta, 1999), and the reading Y-CAIMI scorehas been found to correlate with IQ and reading achievement. In addition, authors of Y-CAIMIwere able to measure children over a course of three years to examine the predictability of YCAIMI and its relationship to CAIMI. Weaknesses of Y-CAIMI include a small validationsample of only 107 children, the full scale is not easy to access, few other research studies haveused this measure, it does not measure other aspects of reading motivation beyond a generalintrinsic motivation to reading, and administration is individual; therefore, it will take time toadminister Y-CAIMI to an entire class.Young Reader Motivation Questionnaire. Coddington and Guthrie (2009) developed ascale of reading motivation for early elementary school students named YRMQ in order toexamine motivation of emergent readers and the role gender may play in reading motivation. Thescale was developed using responses from first grade students. The full scale is available inCoddington and Guthrie (2009). A description of its development can be found in Coddingtonand Guthrie (2009). The information in the following two sections was obtained from this

13development paper.Constructs Measured. The 12 item YRMQ (α .70) includes four items measuring selfefficacy for reading (Are you good at remembering words?, α .64), three items measuringreading orientation (Is it fun for you to read books?, α .60), and five items measuringperceptions of difficulty in reading (Do you make lots of mistakes in reading?, α .67). Theauthors used a question format instead of a declarative format since they believed it would beless confusing for younger children. The response to each item is on a 4-point scale (1 No,Never; 2 No, Not Usually; 3 Yes, Usually; 4 Yes, Always). For example, if a student wasasked, “Can you work out hard words by yourself when you read?” the student would beprompted to reply with a yes or no. If the student answered yes, he or she would be asked, “Canyou work out hard words by yourself always or usually?” If the student answered no, he or shewould be asked, “Can you not usually work out hard words by yourself or never?” These itemswere given individually and orally to each student.Development and Validation. The measure was field-tested with 84 first grade studentsfrom two mid-Atlantic elementary schools. To measure construct validity the research teamcompared YMRQ scores to measures of reading achievement (Woodcock-Johnson Letter-WordIdentification subtest) and measures of reading motivation with a teacher form, T-YMRQ, whichasked similar questions to those included on YMRQ. Self-efficacy correlated significantly andpositively to word identification and perceptions of difficulty correlated significantly andnegatively to word identification. The student responses for all three sub-scores correlatedsignificantly with the matching teachers sub scores on T-YMRQ. Students’ perceptions ofdifficulty were significantly related to both students’ self-efficacy (-.50) and reading orientation(-.36). Orientation however was not significantly correlated with self-efficacy. When examining

14these relationships within gender, difficulty was still significantly related to self-efficacy (-.68)and orientation (-.54). However, none of the constructs were correlated when examining scoresfrom girls. In general, girls were statistically more motivated than boys. When statisticallycontrolling for the other motivation constructs, both efficacy and difficulty in reading statisticallypredicted word-identification scores in multiple regressions. When these analyses were runseparately for each gender, only efficacy predicted word-identification with boys.Critique. Overall, YMRQ is a fairly reliable, valid scale of motivation of emergentreaders. Strengths of YMRQ include that it has been translated to Chinese (Wang & Coddington,2014), its developmentally appropriate format, it is easy to access, and findings indicating itsrelationship to word-identification scores and teacher ratings of similar constructs. Despite thesestrengths there are several weaknesses of YMRQ. First, despite being a measure of readingmotivation, most of the subcontracts seem to be similar to items found on scales of readingefficacy and self-concept (Tunmer & Chapman, 1991). In addition, reliability scores of thesubscales were all below .70, the validation sample was only a small sample of first gradechildren which might not be generalizable, no exploratory or confirmatory factor analyses wereconducted, and the scale is individually administered which may take a while if a teacher wantedto measure his or her whole class of students.Reading Motivation Scales: ElementarySeven reading motivation scales fell best into an elementary category. These are CAIMI,ERAS, Motivation to Read Profile (MRP), Motivation for Reading Questionnaire (MRQ),Reader Self-Perception Scale (RSPS), Reading Self-Concept Scale (RSCS), and SRQ-ReadingMotivation (SRQ-RM) Questionnaire. A short review of each scale is included.Children’s Academic Intrinsic Motivation Inventory. CAIMI (Gottfried, 1985) was

15developed to measure intrinsic motivation for reading, math, social studies, and science.According to Gottfried (1985) “Intrinsic motivation concerns the performance of activities fortheir own sake in which pleasure is inherent in the activity itself” (p. 631). CAIMI measuresaspects of intrinsic motivation such as “enjoyment of learning, an orientation toward mastery,curiosity, persistence, task endogeny, and learning of challenging, difficult, and novel tasks” (p.633). CAIMI has been assessed with students in grades four through eight. Sample items and adescription of its development are available in Gottfried (1985). The information in the followingtwo sections was obtained from this development paper.Constructs Measured. The 122 item CAIMI (α .71-.92) includes 26 items measuringintrinsic motivation for reading (α .71), 26 items measuring intrinsic motivation for math(sample, α .71), 26 items measuring intrinsic motivation for social studies (α .73), 26 itemsmeasuring intrinsic motivation for science (α .69), and 18 items measuring intrinsic motivationfor general school learning (α .67). Of the 112 items, the authors wrote 28 items specificallyfor reading, including items measuring high motivation (I enjoy learning new things in reading.),low motivation (New ideas are not interesting to me in reading.), and items requiring forcedchoice (Is it more important to you to do a school assignment to learn more or get a good gradein reading?) (Gottfried, 1985, p. 634). Students respond on a “5-point Likert scale ranging fromstrongly agree (1) to strongly disagree (5)” for the high and low motivation items (p. 633). Ascale administer reads the instructions and items on the scale aloud to the child. This scale is hasbeen both individually and group administered.Development and Validation. Gottfried (1985) presented three separate studies exploringthe development of this scale. Study 1 included 141 fourth and seventh grade suburban schoolchildren. Study 2 included 260 children in grades 4-7 in a middle class neighborhood. Study 3

16included 166 middle class students in grades 5-8. In Studies 1 and 3, CAIMI reading motivationcorrelated significantly with reading and language achievement tests. In Study 3, CAIMI readingmotivation was correlated significantly with grades in reading, social studies, and science. InStudies 1 and 2, CAIMI reading motivation decreased as student age increased and readingmotivation correlated positively to perceptions of competence. In Study 3, CAIMI readingmotivation was correlated significantly with the three scales from Harter’s (1981) Scale ofIntrinsic versus Extrinsic Orientation in the Classroom. In all three studies, CAIMI readingmotivation correlated negatively and significantly with reading anxiety.Critique. Overall, CAIMI is a reliable, valid scale of intrinsic motivation for upperelementary and middle school students in the content areas of reading, math, social studies, andscience. Strengths of CAIMI include having a significant relationship to reading and languageachievement, grades in reading, reading anxiety, and scales from Harter’s (1981) Scale ofIntrinsic versus Extrinsic Orientation in the Classroom. The reading subscale of CAIMI wasabove .70, can be individually or group administered, and has been used in research on topicssuch as gifted students’ motivation (Gottfried & Gottfried, 1996; Gottfried, Gottfried, Cook, &Morris, 2005), relations of reading motivation to achievement, IQ, and perceptions ofcompetence (Gottfried, 1990), learning disabilities (Wilson & David, 1994), relations ofmotivation to anxiety (Gottfried, 1982), student-centered environments (Mathews, 1991), andremediation of at-risk boys (Rawson, 1992).Despite these strengths there are several weaknesses of CAIMI including that it onlymeasures intrinsic motivation to read and does not measure other constructs of readingmotivation, it was validated with a small sample of middle class students and may not generalizeto students from other schools and backgrounds, no exploratory or confirmatory factor analyses

17were conducted, and the full scale is not easily accessible.Elementary Reading Attitude Survey. ERAS was developed to fill a need for a publicdomain quantitative group survey with desirable psychometric attributes (McKenna & Kear,1990). The authors also wanted a survey that teachers could use to estimate attitudes aboutreading efficiently and reliability. The survey was written based on the historical research onattitude and achievement. Items were developed to tap children’s attitudes towards recre

motivation scales for their research and classroom. The goals of this review were to identify a set of reading motivation student self-report scales used in research, examine the development and psychometric properties of each reading motivation scale, and compare scales on availability, reliability, age range, and motivation constructs measured.