5
659 JOURNAL OF APPLIED BEHAVIOR ANALYSIS 1998, 31, 659–663 NUMBER 4(WINTER 1998) EFFECTS OF CONTINGENT REWARD AND INSTRUCTION ON ORAL READING PERFORMANCE AT DIFFERING LEVELS OF PASSAGE DIFFICULTY GEORGE H. NOELL,KRISTIN A. GANSLE,JOSEPH C. WITT, ERNEST L. WHITMARSH,JENNIFER T. FREELAND,LYNN H. LAFLEUR, DONNA N. GILBERTSON, AND JOHN NORTHUP LOUISIANA STATE UNIVERSITY This study examined the effects of reinforcement contingencies designed to increase the performance of existing reading skills as well as the effects of instruction—modeling and practice—designed to increase skill level for oral reading fluency across three levels of reading materials. Results showed that a combination of contingencies, modeling, and practice was effective in producing substantial increases in reading fluency for all partic- ipants at their assigned grade levels. These results demonstrate one strategy for experi- mentally determining those instructional components that are required to increase oral reading rate. DESCRIPTORS: oral reading, skill and performance, task difficulty The analysis of functional relationships between instructional variables and oral reading rate has the potential for extending the application of behavior analysis to a high-incidence problem that is socially im- portant. Recent work within applied behav- ior analysis has demonstrated the utility of analyses of antecedent instructional variables (e.g., Daly & Martens, 1994). Previous lit- erature has also examined the effect of re- inforcing consequences on academic behav- iors (Pereira & Winton, 1991). Reinforced practice can contribute to developing fluen- cy and may also increase the performance of skills that are already in the individual’s rep- ertoire. The integration of instructional events that maximize current performance and events that produce acquisition of new skills into a single analysis of oral reading fluency may contribute to a more systematic approach to instructional decision making. The current study was designed to examine a strategy for assessing the separate and com- Correspondence should be directed to George Noell, Department of Psychology, Louisiana State University, Baton Rouge, Louisiana 70803-5501 (E- mail: [email protected]). bined effects of contingencies and instruc- tion on oral reading fluency. METHOD Participants, Setting, and Materials Participants in this study were 3 boys (all 9 years old) enrolled in a 3-week summer program for children with a diagnosis of at- tention deficit hyperactivity disorder. The participants were beginning fourth-grade students who were of at least average intel- lectual functioning and whose parents had expressed concern regarding their reading skills. Each participant took the same med- ication (methylphenidate) throughout the study; however, Mike’s dosage changed dur- ing the study. Sessions were conducted in the classroom with students seated away from other students. Twenty-one reading passages at each grade level were selected randomly from Harcourt Brace Jovanovich Treasury of Literature, em- ploying standard curriculum-based measure- ment guidelines for the selection of passages (see Shapiro, 1989). The mean Flesch-Kin- caid readability index for the grade levels in-

Effects of contingent reward and instruction on oral reading performance at differing levels of passage difficulty

  • Upload
    j

  • View
    213

  • Download
    1

Embed Size (px)

Citation preview

Page 1: Effects of contingent reward and instruction on oral reading performance at differing levels of passage difficulty

659

JOURNAL OF APPLIED BEHAVIOR ANALYSIS 1998, 31, 659–663 NUMBER 4 (WINTER 1998)

EFFECTS OF CONTINGENT REWARD ANDINSTRUCTION ON ORAL READING PERFORMANCE AT

DIFFERING LEVELS OF PASSAGE DIFFICULTY

GEORGE H. NOELL, KRISTIN A. GANSLE, JOSEPH C. WITT,ERNEST L. WHITMARSH, JENNIFER T. FREELAND, LYNN H. LAFLEUR,

DONNA N. GILBERTSON, AND JOHN NORTHUP

LOUISIANA STATE UNIVERSITY

This study examined the effects of reinforcement contingencies designed to increase theperformance of existing reading skills as well as the effects of instruction—modeling andpractice—designed to increase skill level for oral reading fluency across three levels ofreading materials. Results showed that a combination of contingencies, modeling, andpractice was effective in producing substantial increases in reading fluency for all partic-ipants at their assigned grade levels. These results demonstrate one strategy for experi-mentally determining those instructional components that are required to increase oralreading rate.

DESCRIPTORS: oral reading, skill and performance, task difficulty

The analysis of functional relationshipsbetween instructional variables and oralreading rate has the potential for extendingthe application of behavior analysis to ahigh-incidence problem that is socially im-portant. Recent work within applied behav-ior analysis has demonstrated the utility ofanalyses of antecedent instructional variables(e.g., Daly & Martens, 1994). Previous lit-erature has also examined the effect of re-inforcing consequences on academic behav-iors (Pereira & Winton, 1991). Reinforcedpractice can contribute to developing fluen-cy and may also increase the performance ofskills that are already in the individual’s rep-ertoire. The integration of instructionalevents that maximize current performanceand events that produce acquisition of newskills into a single analysis of oral readingfluency may contribute to a more systematicapproach to instructional decision making.The current study was designed to examinea strategy for assessing the separate and com-

Correspondence should be directed to GeorgeNoell, Department of Psychology, Louisiana StateUniversity, Baton Rouge, Louisiana 70803-5501 (E-mail: [email protected]).

bined effects of contingencies and instruc-tion on oral reading fluency.

METHOD

Participants, Setting, and Materials

Participants in this study were 3 boys (all9 years old) enrolled in a 3-week summerprogram for children with a diagnosis of at-tention deficit hyperactivity disorder. Theparticipants were beginning fourth-gradestudents who were of at least average intel-lectual functioning and whose parents hadexpressed concern regarding their readingskills. Each participant took the same med-ication (methylphenidate) throughout thestudy; however, Mike’s dosage changed dur-ing the study. Sessions were conducted inthe classroom with students seated awayfrom other students.

Twenty-one reading passages at each gradelevel were selected randomly from HarcourtBrace Jovanovich Treasury of Literature, em-ploying standard curriculum-based measure-ment guidelines for the selection of passages(see Shapiro, 1989). The mean Flesch-Kin-caid readability index for the grade levels in-

Page 2: Effects of contingent reward and instruction on oral reading performance at differing levels of passage difficulty

660 GEORGE H. NOELL et al.

Figure 1. Median number of words read correctly graphed by session, level of reading material, and ex-perimental phase for Glen. The mastery, instructional, and frustrational ranges on all figures are based on Fuchsand Deno (1982).

cluded in this study were second grade, 2.57(range, 2.0 to 2.89); third grade, 3.71(range, 3.11 to 3.94); fourth grade, 4.52(range, 4.1 to 4.96); and fifth grade, 5.48(range, 5.1 to 5.97). The assessment includ-ed reading material from the grade the childwould attend in the fall and the two preced-ing grades for each participant. For Mike,fifth-grade materials were substituted for sec-ond-grade materials based on summer pro-gram records that indicated that his instruc-tional level may have approached fourthgrade. The dependent variable was the me-dian number of words correct per minute

(WCPM) the participant read from threepassages within each session. Interobserveragreement data were obtained for a random-ly selected 33% of sessions for each partici-pant and were calculated based on exactword-by-word agreement. The mean inter-observer agreement was 98.3% and neverfell below 95.6%.

Experimental Design,Procedures, and Reliability

Each participant’s reading was evaluatedusing a multiple baseline design across threelevels of reading materials. Contingent re-

Page 3: Effects of contingent reward and instruction on oral reading performance at differing levels of passage difficulty

661CONTINGENCIES, INSTRUCTION, AND TASK DIFFICULTY

Figure 2. Median number of words read correctly graphed by session, level of reading material, and ex-perimental phase for Mike. Round data points correspond to placebo sessions, triangular data points indicate0.3 mg/kg of methylphenidate, and square data points indicate 0.6 mg/kg of methylphenidate.

ward was applied first, followed by reward,modeling, and practice if WCPM was notin the mastery range. If contingent rewardwas ineffective but reward, modeling, andpractice were effective, then modeling andpractice alone were applied at the next gradelevel. This modification was used to test forthe possibility that modeling and practicewithout reward provided an effective treat-ment. Each session consisted of the partici-pant reading three different randomly se-lected passages from one grade level. Using

a checklist (available from the first author),an observer measured treatment integrity for33% of sessions that were randomly selectedfrom all phases for all participants. Themean treatment integrity was 98.4%.

Baseline. Participants were asked to dotheir best to read the passage aloud and toattempt each word. If a participant did notcomplete a word within 3 s, the experi-menter told the participant the word andmarked it as incorrect. No programmed con-sequences were provided for reading perfor-

Page 4: Effects of contingent reward and instruction on oral reading performance at differing levels of passage difficulty

662 GEORGE H. NOELL et al.

Figure 3. Median number of words read correctly graphed by session, level of reading material, and ex-perimental phase for Jack.

mance, and no feedback was provided toparticipants.

Contingent reward. The participant wasprovided with a token coupon if his medianWCPM exceeded his median WCPM fromthe previous session with no increase in er-rors. Token coupons had previously beenidentified as effective reinforcers for mathe-matics responding for each participant andcould be exchanged for rewards available inthe summer program. Glen (at Session 10)and Mike (at Session 11) were given the op-tion of selecting 50 cents in lieu of a couponbecause of concerns regarding possible sati-

ation based on the large number of tokencoupons they received in other activities inthe summer program.

Modeling and practice. Modeling and prac-tice sessions began with the experimenterreading the passage to the student (listeningpassage preview, Daly & Martens, 1994) toa point 20% beyond the participant’s me-dian score from the previous session. Afterthe student practiced reading the passageonce with the experimenter correcting anyerrors, a timed reading of that passage wasobtained. This sequence was repeated for allthree passages in a session.

Page 5: Effects of contingent reward and instruction on oral reading performance at differing levels of passage difficulty

663CONTINGENCIES, INSTRUCTION, AND TASK DIFFICULTY

Reward, modeling, and practice. This phasecombined the contingent reward, modeling,and practice procedures as described in thepreceding two conditions.

RESULTS AND DISCUSSION

Glen’s second-grade level WCPM (Figure1) increased under contingent reward, andcontingent reward therefore was applied tothird-grade materials. No increase in readingwas observed, so modeling and practice wereadded to the treatment, and third-gradeWCPM increased. These data suggested therelative importance of modeling and prac-tice, so modeling and practice alone were ap-plied to fourth-grade materials. Reward wasadded to modeling and practice when theywere found to be ineffective. Contingent re-ward was ineffective for Mike (Figure 2) inthird-grade materials and was only moder-ately effective for fourth- and fifth-grade ma-terials. The addition of modeling and prac-tice to each level was followed by increasedWCPM. Jack’s reading rate (Figure 3) didnot increase when contingent reward was ap-plied to second- or third-grade materials.Additional instruction was not applied tosecond-grade materials because Jack’s re-sponding fell within the mastery range. Theincrease in WCPM following the addition ofmodeling and practice to third-grade mate-rials indicated the relative importance ofthese treatment components. Based on theseresults, modeling and practice alone were ap-plied to fourth-grade materials and resultedin increased WCPM.

The analyses conducted in this studyidentified procedures that resulted in an in-crease in mean reading rate of 59% or great-er between baseline and the final treatment

phase in the assigned grade level for all par-ticipants. Glen and Mike demonstrated thegreatest increase in reading fluency when re-ward, modeling, and practice were com-bined, but only modeling and practice ap-peared to be critical for Jack.

A number of limitations of this prelimi-nary study can be addressed by future re-search. First, the multiple baseline designmay not be sufficiently flexible to be an op-timal approach to this type of assessment.For example, the decision to introduce mod-eling and instruction alone in two data seriesprecluded direct replication of the precedingdata series. Second, using reading materialswith a wider range of difficulty may haveresulted in greater differentiation amongtreatments. Third, the potential importanceof sequence effects is unknown. Fourth, theexposure to the same passage in two to threesessions over the 3 weeks may have resultedin some passage-specific learning that wasnot the focus of the analysis.

REFERENCESDaly, E. J., III, & Martens, B. K. (1994). A com-

parison of three interventions for increasing oralreading performance: Application of the instruc-tional hierarchy. Journal of Applied Behavior Anal-ysis, 27, 459–469.

Fuchs, L. S., & Deno, S. L. (1982). Developing goalsand objectives for educational programs. Washing-ton, DC: American Association of Colleges forTeacher Education.

Pereira, J. A., & Winton, A. S. W. (1991). Teachingand remediation of mathematics: A review of be-havioral research. Journal of Behavioral Education,1, 5–36.

Shapiro, E. S. (1989). Academic skills problems: Directassessment and intervention. New York: Guilford.

Received January 12, 1998Initial editorial decision March 4, 1998Final acceptance June 23, 1998Action Editor, Brian K. Martens