Scenario

At Madison Middle School, the sixth grade language arts teachers are unhappy with the curriculum. However, there's a disconnect, as standardized test scores show improvement from previous years. You, a district reading specialist, have been called in to investigate the situation. Sixth grade reading scores have improved moderately from 2014 to 2018. These increases amount to 0.23 SD during the five year period for which reading scores were recorded. In general, a 0.50 standard deviation increase represents an increase from the 50th to the 69th percentile for data that are normally distributed. This compares to a percentile increase from the 50th to the 60th when the corresponding increase in standard deviation is approximately 0.25.

Part 1 - Paired Samples t-Test

Sandbox Scenario 1

As you begin to take a closer look at the sixth grade language arts issues, you observe that there are twenty students enrolled in a remedial reading program. Several of the teachers who you surveyed questioned whether this program was effective because this was the program had been handed to them by administration. This point of contention continues to make them disgruntled. Since no quantitative analysis had ever been conducted to determine the reading program's efficacy, you make a decision to design a study so that finally, a determination can be made about whether the reading program is effective. Specifically, you are going to conduct inferential statistics to determine if the remedial reading program actually helped improve reading scores. The students' reading scores before and after the intervention are shown in the dataset provided.

Instructions

1. Using the dataset provided, conduct statistical analysis to test the following null hypothesis:
   • There is no statistically significant difference between students' reading scores before and after the intervention is provided to students.
2. Your results will indicate that you either will reject the null hypothesis when there is a significant difference or you will fail to reject the null hypothesis when there is no significant difference.

Dataset Legend

https://1drv.ms/x/s!AhnxthXizDn8jVI-2JvgDtfvwcBZ?e=UFvzrQ

The following legends are used in the sandbox dataset.

• Gender: 0 = Male; 1 = Female.
• Race: 1 = Asian/Pacific Islander; 2 = African American; 3 = Hispanic/Latino; 4 = White/Caucasian.
• Socioeconomic status (SES): 1 = 75,000-100,000; 2 = 50,000-69,999; 3 = 20,000-49,999.
• School type: School 1 and School 2.
• Academic program: 1 = program 1; 2 = program 2; 3 = program 3