of students on the disability service program on Irvine University
A Statistical Approach
Reading is one of the first abilities developed and enforced during the first years of education across the world. Duggan (2009) states that one of the two most characteristic attributes a skillful reader has is his speed and ease while reading, the other being the understanding of the text. Conversely, the experience of reading can be enhanced or deteriorated by outside elements other than the reader himself. Irlen’s study (as cited by Morrison, 2011) suggests that visual perception is distorted when reading black words on a white background, due to the high contrast. Morrison (2011) continues this idea, by also explaining that the brightness of the background affects the ability to distinguish the black letters.
The population for this particular study focuses on the students of the disability services program of the Irvine University that could benefit from using colored filters to improve their reading experience. The sample consists on 21 of these students, randomly selected, involving a group of varying ages, majors, abilities, majors and socioeconomic factors.
Research Question
1. What are the differences (if any) in time between reading a passage in light blue-colored paper as compared to a different, but similar passage on white paper among disability services program students of Irvine University?
Receiving a positive answer on this matter (results showing that students are able to read light blue-colored paper faster than traditional white paper) would allow Irvine University to invest on materials and strategies to improve the quality of their alumni’s reading comprehension experience. Otherwise, having a negative or neutral result would prevent the institution from making an unfruitful investment.
Independent Variable
Paper Color
Dependent Variable
Reading Time
The identified variables used in the experiment were (1) Paper Color and (2) Reading Time. Paper Color is a nominal variable, as “White Paper” and “Light Blue-Colored Paper” are two separate categories without any ordering, while Reading Time is a ratio variable, taken in seconds. The sampling methodology involved randomly testing 21 students in their learning disability program, during the beginning of the Fall period in 2015. These students had different major, socioeconomic status and abilities, as well as varying ages. On the first day of the test, students were asked to read a passage on white paper, then to return the next day to read a different (but similar) passage of identical length and difficulty level on a light blue-colored paper. Measures were taken upon how long they spent reading each of the passages.
After identifying the method, there are a few variables that are not present in the study that could have an impact on the results.
Confounding Variables
Text’s Length
Text’s Difficulty Level
Lurking Variables
Text’s Topic
Room’s Illumination
The original study controls the length and difficulty of the passages, as White Paper texts and Light Blue-Colored texts have equal levels of these indicators, making sure the Reading Time variable is not affected by differences between them, if these were not constant, the would have undesired effect on the study. However, the text’s topic and student’s major were not taken into account, allowing a wide margin of uncertainty that these two variables could introduce. For example, a Psychology major would have an easier time reading a humanistic passage on Whiter Paper and then experience some difficulty if the Light Blue-Colored Paper text’s topic was related to Statistics. Also, as all White Paper tests were made on day 1, and then all Light Blue-Colored tests were evaluated on day 2, there is no way of knowing whether the illumination on both days was the same, as this variable wasn’t taken into account. These three variables could have had an undesired impact on the study.
This is an experimental and quantitative study, as students are subject to a test that generates results measured in seconds (a unit of time) and raw data is collected (Kumar, 2014).
Morrison (2011) used color overlays for her study and set three nominal levels: clear, optimum and random. In her experiments, she actually found that participants didn’t improve their reading time by using the optimum overlay, suggesting that color filters may not be a justified investment.
A bar graph was used to compare both scores on Reading Time on both types of Papers. A quick evaluation of the table reveals that reading on the Light Blue-Colored Paper took students a few more seconds in most cases. Only one student had a faster reading on White Paper (student 10). This is considered a possible outlier.
Graph 1. Reading Time of Each Student for Both Types of Papers.
Since the experiment recorded a small amount of samples (only 21 students were tested), scores were then grouped to see the frequency of scores and detect any abnormality within the results. The groups, which stored the scores every ten seconds were set up and both types of paper were contrasted against each other in both a Histogram and a Frequency Polygon. Although not perfectly bell-shaped (slightly skewed), the frequency polygon showed the normality was not violated within the study. Most students spent between 60 and 90 seconds reading on both types of paper.
Graph 2. Frequency of Scores based on groups of tens for Reading Time.
Calculating variance, standard deviation and range revealed that both tests are somewhat scattered, but Light Blue-Colored Paper tests showed to be more disperse than Whiter Paper tests.
After that, outliers were studied within the sample. The 1.5 IQR was used as the suggested method by (Kumar, 2014). For this process, the first and third quartiles for each of the samples was calculated to find the inner fence, looking for Minor Outliers. For the Light Blue-Colored Paper tests, the inner fence value was between 25 and 145 seconds, none of the test results for this category yielded such result. Moreover, the inner fence for White Paper results was between 35 and 119. There’s only one value that slightly strayed from the inner fence, which was Student 21’s result for White Paper results with 120 seconds. No corrections were made for this study, since the only outlying value for this experiment is a fairly mild one, therefore, no implications would occur by using the current data.
After taking into account the descriptive data, some of the facts began hinting some ideas about the research. In most cases, it’s fair to assume that students had a more difficult time when reading on a Light Blue-Colored paper, since the results reflected a longer time executing this action.
Null Hypothesis
Irvine students in the disability service program take the same amount of time when reading a passage in a White Paper text than in a Light Blue-Colored text.
H0: Difference between Whiter Paper Text Reading Time and Light Blue-Colored Text Reading Time is 0.
Alternative Hypothesis
Irvine students in the disability service program take more time when reading a passage from the “innovative” Light-Blue Colored text, than with the White Paper text.
H1: Difference between Light Blue-Colored Text Reading Time and White Paper Text Reading Time is greater than 0.
As mentioned above, because the interest of this test is heavily influenced on whether or not the difference between Light Blue-Colored Text Reading Time and White Paper Text Reading Time is greater than 0 and not otherwise because of the accumulative data, this test is a one-tailed test. The chosen significance level is 5% (Kamur, 2014)
Applying a Hypothesis test, it’s clear that for Hypothesis 0, the mean of a normalized distribution is 0 seconds (It takes the same amount of time to read both types of paper). Performing a T-Score test will test the hypothesis:
t=x-μ0s/Sample Size=8.33-05.60/21=6.81
degrees of freedom=n-1=21-1=20
Taking the t-value and degrees of freedom, there’s a 0% chance that given the test results Hypothesis 0 is true. This is confirmed in the scatter plotted showed below, where there’s a strong indication that readers spent more time on passages written on Light Blue Colored paper. All in all, the data collected in Part I was clearly interpretable on Part II, little by little becoming more evident that the null hypothesis was wrong. Without the data collection and the right methods, inferential statistics wouldn’t be able to deliver any relevant data.
It’s important to note that both descriptive and inferential statistics based their results on the data collected in the beginning. Thus, collecting the wrong information could end up in a study that delivers a far-fetched result. Nevertheless, even though the data collected was useful, taking into account the variables noted in PART I would probably have changed the results.
Information about age, relationship between majors and the text’s content would’ve made the study more accurate, as they would’ve eliminated any uncertainty about the ability of the reader to comprehend the text, which as Irlen’s (as cited by Morrison, 2011) study suggested, is one of the indicators of a good reader. Also, controlling the environment of the test: lighting conditions and randomizing the order of the readings (instead of choosing to do all tests for Whiter Paper on one day, and all tests for Light Blue-Colored Paper on the other) is a necessary measure that future researchers must conduct.
References
Duggan, Mauna Sue. (2009) A Critical Examination of the Effects of Colored Paper on the Academic Achievement of Fourth Graders in Reading Comprehension and Vocabulary (Unpublished Dissertation). Auburn University, Auburn, Alabama.
Morrison, Rhonda F. (2011) The Effect of Color Overlays on Reading Efficiency (Doctoral Dissertation). Retrieved from http://scholarworks.umass.edu/cgi/viewcontent.cgi?article=1432&context=open_access_dissertations.
Kumar, Ranjit. (2014) Research Methodology, fourth edition. New York, NY: Sage Publications.
Appendix A
Original Exercise:
Study #3: Psychology
Some people believe that reading text printed on colored paper is easier than reading text on white paper. For this reason, disability services locations often provide colored filters that are clear to put over top of books and imprinted pages. Irvine University, a public research institution with 30,000 undergraduate students is considering the possibility of providing colored filters or materials on colored paper for those people in disability services who would be helped by this tool. Irvine University randomly tested 21 students in their learning disability program during the first week of classes in Fall 2015. The students, of varying ages, abilities, majors, and socioeconomic status were timed as they read a passage printed in black ink printed on white paper and timed reading the same passage printed with black ink on light blue-colored paper. Students were tested on one day with a passage on white paper and then asked to willfully return the next day to read a similar, but different passage of the same length and difficulty level on light blue-colored paper. Is there sufficient evidence to show that reading text on light blue-color paper is more effective? You are asked to analyze this data and present your
results to the company with a hypothesis test on whether or not there is significant improvement, a correlation between the two variables, and a regression analysis. Also include in your report research from journals, articles, and national studies that show whether or not there is a relationship between colored paper and improvement in reading.
