Design, reliability and validity of a paediatric rate of reading (PRR) chart.

Abstract

Background: Reading rate is a measure of fluency, reflecting the level of reading performance especially in children, which is not typically measured during routine eye examinations. Optometric clinical tests such as Snellen visual acuity are often poor predictors of everyday reading performance, as they test the smallest print a person is able to read rather than fluency. Conventional reading rate tests for educational purposes presents with many limitations; they concentrate on linguistic skills, increase in complexity as the reading progresses, limited by the readers vocabulary but more importantly, they do not take the level of the child's vision into consideration. There is currently no reading rate chart that is designed with optometric notations specifically for children with normal vision and low vision. It is therefore necessary to design a reading rate chart that takes the above limitations into consideration. Aim: This study aimed to design a chart that can be used to measure reading rates in normal sighted and low vision primary school children. Methods: The aim of the study was achieved in four parts; the design, reliability, validity of a reading rate chart and finally the testing of the chart on low vision participants. In the design of the chart, ten frequently used words in grade one English reading books were randomly selected from five primary schools in KwaZulu-Natal province, South Africa. The reliability and validity of the chart were established on normal sighted children, aged nine to twelve years from two primary schools in the Durban area chosen by convenience sampling method, with sample size of 100 for reliability and 100 for validity. Reliability was established with test and retest reading rates using the new chart while validity was established by determining the reading rates using new the chart and the Wilkins reading rate chart. Data were analyzed using the Paired t-test, Pearson correlation, and Bland and Altman method. Finally, the testing of the new chart without and with low vision device, on fourteen low vision children, aged eight to nineteen years, attending a school for the visually impaired in KwaZulu-Natal. Data was analyzed using Paired t-test and Pearson correlation. Results: The words were arranged in random order, ten words per row and ten rows per paragraph. The chart consisted of six paragraphs (versions A, B, C, D, E and F) with six acuity levels and four optometric notations. Each version was printed on a separate sheet, in Arial and Times New Romans fonts and printed in black ink on approximately white cardboards. In reliability, the mean test and retest reading rates were 77.65 ± 25.30 and 78.23 ± 24.70 (p = 0.29, R² = 0.95). In Bland and Altman method, the mean difference was −0.58 with confidence limits at +10.07 and -11.23. In validity, reading rate for Wilkins chart and the new chart were 75.82 ± 23.64 and 74.92 ± 23.58 (p = 0.01, R² = 0.99) respectively. In Bland-Altman method, the mean difference was +0.90, upper limit at +6.33 and lower limit at –4.53. The mean reading rate, of the low vision children, without and with the low vision device were 59.32 ± 24.08 words per minute (wpm) and 67.04 ± 25.63 words per minute (wpm) respectively (p = 0.09 and r = 0.82). Conclusions: This chart can be used for reading rate assessment for both normally sighted and low vision children and is statistically reliable and valid.