print this page

Comparison of Two Error Correction Procedures Within a Precision Teaching Framework
During Learning to Read Words

Angela Galvin, Michelle I. Harrington, Edward Langford, Matthew L. Israel Ph.D.

Judge Rotenberg Educational Center

 240 Turnpike St.
Canton, MA 02021
(781) 828-2202

Based on a poster presented at Association for Behavior Analysis Conference, May 2005, Chicago, IL

1)      Introduction

The Judge Rotenberg Educational Center (www.judgerc.org) operates day and residential programs for children and adults with behavior problems, including conduct disorders, emotional problems, brain injury or psychosis, autism, and developmental disabilities.  The fundamental approach taken at JRC is the use of behavioral psychology and its various technological applications, including behavioral education, programmed instruction, precision teaching, behavior modification, behavior therapy, behavioral counseling, self-management of behavior, and chart-sharing.

At the Judge Rotenberg Center (JRC) the implementation of a significant portion of the reading curriculum is based on the principles of Precision Teaching.  This includes conducting multiple assessment timings with students on the reading of words until they reach a high rate of correct responses and a low rate of incorrect responses per minute within each portion of the curriculum. 

This study examined how two different types of error correction methods affected learning and retention of the proper pronunciation of words presented in a specific reading task. Curriculum material was selected from the series, Teach Your Children to Read Well and was presented in one minute time samples. Students were given two different word lists that were at their reading decoding level.  During timings for the first word list, students were corrected during each timing, as an error occurred.  During timings for the second word list, students were corrected at the end of each time sample. Students received the same verbal correction procedure for each word list; the only difference was when the correction occurred. Subjects were similar in age.  All subjects were tested on the material one week after mastery had been achieved to determine how their retention differed. 

2)      Method

This study was comprised of 9 students, all of whom were enrolled at JRC.  There were 6 males and 3 females.  The students had an average age of 16.3 years.  The students’ IQ levels ranged from 44 to 97.  The average reading grade equivalence (G.E.) was 4.5.  Each of these students was functioning far below his or her same-aged peers whose average grade equivalence was 10.3. Students had diagnostic labels that include learning disorders, disruptive behavior disorders, mood disorders, anxiety disorders, pervasive developmental disorders, organic brain syndrome and mental retardation.  All participants lived in JRC group homes and attended school at JRC’s main school building. 

A trained educational staff worked with each student for five days completing 6 timings per day.  During these daily sessions, participants completed three one-minute timings on Word List 1 (errors corrected during the timing) and three one-minute timings on Word List 2 (errors corrected after the timing).  During each session, the students were given Word List 1 first.  For all timings, the educational staff indicated to the student where to begin on the page, set the timer to 1-minute and said, “Please begin.” The student read the words from left to right or top to bottom, depending upon the starting point.  If the student skipped, or mispronounced a word, during Word List 1 timings, the educational staff immediately provided corrective feedback and prompted the student to continue with the list until the timer stopped.  If the student skipped a word or mispronounced a word, during the Word List 2 timings the educational staff made a note of this and did not indicate to the student that an error was made.  The student continued with the list until the timer stopped. After the timings were completed, the educational staff reviewed the mispronounced words with the student.  Data for time samples from both word lists were plotted on the Standard Celeration Chart, immediately following each timing (Lindsley, 1992a, p. 51).  

3)      Results

For Word List 1, the participants’ average reading rate started at 50 words correct and 9 incorrect per minute and increased to 132 words correct and 2 words incorrect per minute.  For Word List 2, the participants’ average reading rate started at 55 words correct and 10 words incorrect per minute and increased to 121 words correct and 2 incorrect per minute. 

Data are displayed on three Daily Standard Celeration Charts.  Each chart has data for three students’ best per day timing of each session during Word List 1 and Word List 2.  The best per day rate was figured by the highest correct and lowest incorrect response.  Figure 1, Figure 2 and Figure 3 show data for students who had IQs in the moderate MR range; the mild MR range and the borderline-average range, respectively.

The data for the difference in rates correct and incorrect for both word lists between the first and last timings are shown on a Weekly Standard Celeration Chart in Figure 4. The data for the rates correct and incorrect for the first and last timings for both word lists are shown on a Daily Standard Celeration Chart in Figure 5.

Two weeks after mastery had been achieved, the students were post-tested.  The post-test timings on both Word List 1 and Word List 2 showed retention and maintenance of the learned words.  The students had to reach their own fluency aim at the end of the study.  The students maintained their fluency rate during the post-test as well. 

4) Discussion

Figures 1, 2, and 3 show that students 1, 2, 3, 4, and 5 produced higher celeration in their corrects during Word List 1.  As shown by the Figure 4 chart, seven students had higher rates of corrects during timings for Word List 1.  Five of those seven students had higher rates of correct responses and lower rates of incorrect responses during Word List 1. Data suggests that students who have Full Scale IQs of 61 and below benefit from being corrected during the timings and students with Full Scale IQs of 66 and above benefit from being corrected after the timing.  Further study will examine the effects of the different error correction procedures on reading selections that are at a higher level of difficulty for the subjects.