Description:

 

The State of Texas Assessments of Academic Readiness, commonly referred to as its acronym STAAR, is a series of standardized tests used in Texas public primary and secondary schools to assess a student’s achievements and knowledge learned in the grade level. Starting in third grade, Texas students are required to take the State of Texas Assessment of Academic Readiness test, or STAAR, every year.  This dataset contains test scores and demographic information for randomly selected third grade students from the Texoma region

 

This data set contains only data from randomly selected third-grade students in from the Texoma region in the 2023-2024 academic year. The data were collected for the purpose of identifying significant predictors in both the home environment and the scholastic setting that may influence educational outcomes. Of particular interest in this study is the evaluation of a new Prep Course designed to prepare children to be able to navigate and succeed on standardized tests.  This course was introduced to select schools on a trial basis to determine the efficacy of the program.  Also of interest is the family socioeconomic status, indicated by the highest education level of either parent (or guardian) and the indication whether the child qualifies for free or discounted lunch.  Test scores are percentile ranks relative to the statewide performance of third grade students in the three facets of the STAAR test: math, reading, and writing.  The following measures were included in the data:

 

Variables:

 

• StudentID– Participant’s identification number which looks strangely similar to the original row number of the dataset.
• Gender –The biological sex of the child, i.e. Female/Male.
• Testing Location – Represents the Independent School District in which the Prep Course and Testing were given Anna, TX and McKinney, TX
• Parental Level of Education –the highest education level completed by either parent (or guardian), categorized by: High School, Associates Degree, or Bachelors Degree.
• Lunch –Binary measure indicating whether or not the child is eligible for free or discounted lunch, indicated by free/reduced for qualifying students and standard for students paying the standard fees.
• Test Prep Course–Binary measure to indicate the completion of the newly introduced Prep Course before STAAR testing began in Spring 2024.  This measure is indicated as Complete for those whose schools were included in the initiative in time to complete the program or None if the student’s school was not included.

 

 

 

Research Question:

 

In this case we will be designing a study to determine if math test scores are significantly improved depending on whether or not the child completed the Test Prep Course, and if the effect differs based on economic status.  Upon loading, examining, and cleaning your data, choose the correct analytical technique to test the following hypotheses:

where TespPrep is the population of kids who have completed the Prep Course, Reduced is the population of kids who are eligible for a Free or Reduced Price Lunch, and Test Prep*Reduced is the interaction between the two.

 

(You may use or quickly recreate the same Frequency and Descriptive Sections from Case 1 to satisfy Accuracy and Outliers if you easily remember what you did to clean the data.)

 

Accuracy:

1. Check the data for out of range scores.
   1. Include a summary showing you do/do not have out of range scores.
   2. If necessary, fix the out of range scores.
      1. Indicate what the problems were in the dataset.
      2. Make all out of range values NA.

• Include a summary showing that you fixed the accuracy issues.

2. Fix the factored columns to have nice labels (i.e. Proper Case, Fully Spelled out). Only factor the IV, do not factor the DV.
   1. Use the data editor in JASP to change the Test Location so that the label for Group A shows Anna, TX and the label for Group B shows McKinney, TX.
   2. Use the data editor in JASP to change the Parent Education variable so that the value for High School, Associate, and Bachelor reads 1, 2, and 3.  This will ensure they line up correctly in your graphs.

Missing data:

3. Exclude all missing data using leastwise deletion.  Use the filter function to eliminate any cases with missing data.

 

Outliers:

4. Use the boxplot feature to show that you have no outliers.
   1. Include a summary of those z-scores.
   2. Do you have any outliers?
   3. Exclude those outliers.

Normality:

5. Perform the proper test to show that the assumption of normality is met.

 

Linearity:

6. Include the multivariate QQ plot.

Homogeneity:

7. Include the multivariate residuals plot.
8. Interpret the graph. Does it indicate homogeneity?

Power:

9. Calculate the number of participants you would need for this study, assuming a medium effect size.
   1. Include a screen shot or summary of the numbers you typed into G*Power, so we can give you partial credit if you get a different sample size than us.

 

ANOVA and Levene’s:

10. Include the ANOVA and Levene’s test output.
11. Do you meet the homogeneity assumption given the results from Levene’s test?
12. Was the overall test significant?
    1. Include the APA/AMA style write up for F (just the statistics):

 

Post Hocs:

13. Calculate the means, standard deviations, and group sizes for your levels.
14. Post hocs:
    1. What type of post hoc testdid you run?
    2. What type of post hoc correction did you run?
    3. Include the t-test output.
15. Effect size:
    1. Calculate the effect size for your pairwise comparisons.
    2. Include the effect size output or MOTE screen shot.
16. Fill in the table below with the information from the above calculations (like the one from the notes):

 

Mean 1	Mean 2	P-value	Explain?	Effect size

 

 

Graph:

17. Include a graph of the means and confidence interval for your ANOVA. Be sure to check the following:
    1. X-axis label
    2. Y-axis label
    3. X-axis group labels
    4. Error bars
    5. Cleaned up graph (no gray backgrounds)

Write up:

18. Write up an analysis of what you find in this data, including all the information you answered above. Use the example in the notes for a guide. This write up should include the following for credit:
19. Result section style (APA and AMA):
    1. Double space
    2. Times New Roman 12 point
    3. Two decimals
    4. Centered, bolded Results
20. Short description of the study/variables.
21. Data screening summary:
    1. Accuracy – did you have problems?  What did you do to fix it?
    2. Missing data – did you have problems?  What did you do to fix it?
    3. Outliers – did you have problems?  What did you do to fix it?
    4. Assumptions:
       1. Normality
       2. Linearity

• Homogeneity and Levene’s

22. ANOVA
    1. Overall F statistic
    2. Post hoc tests / corrections and results
    3. Effect size for all tests
23. Graph with reference to the figure in the text.
    

    Struggling with where to start this assignment? Follow this guide to tackle your assignment easily!

    Step-by-Step Guide to Structuring Your Analysis

    1. Data Cleaning and Preparation

       • Accuracy Check: Review the dataset for out-of-range scores in the math percentile ranks. Any values outside the 0–100 range should be set to ‘NA’. Document any anomalies found and corrected.

       • Labeling Factors: Ensure categorical variables are properly labeled. For instance, update ‘Testing Location’ to reflect ‘Anna, TX’ and ‘McKinney, TX’. Recode ‘Parental Level of Education’ numerically as 1 (High School), 2 (Associate Degree), and 3 (Bachelor’s Degree) for consistency in analysis.

       • Handling Missing Data: Apply listwise deletion to exclude any cases with missing data, ensuring a complete dataset for analysis.

    2. Exploratory Data Analysis

       • Descriptive Statistics: Generate frequency tables and descriptive statistics for all variables to understand the distribution and central tendencies.

       • Outlier Detection: Utilize boxplots and calculate z-scores to identify any outliers in the math scores. Exclude any data points with z-scores exceeding ±3.

    3. Assumption Testing for ANOVA

       • Normality: Conduct the Shapiro-Wilk test to assess the normality of math scores. A p-value greater than 0.05 indicates normal distribution.

       • Linearity: Create a multivariate Q-Q plot to verify the linearity assumption between independent variables and the dependent variable.

       • Homogeneity of Variance: Perform Levene’s Test to check for equal variances across groups. A non-significant result (p > 0.05) supports the assumption.

    4. Power Analysis

       • Use G*Power to calculate the required sample size for detecting a medium effect size (f = 0.25) with an alpha level of 0.05 and power of 0.80. Document the input parameters and the resulting sample size.

    5. Conducting Two-Way ANOVA

       • Model Specification: Set up a two-way ANOVA with ‘Test Prep Course’ and ‘Lunch Status’ as independent variables, and math scores as the dependent variable. Include the interaction term to assess combined effects.

       • ANOVA Output: Report the F-statistics, degrees of freedom, and p-values for each main effect and the interaction.

       • Levene’s Test Result: Interpret the outcome to confirm if the homogeneity of variance assumption holds.

    6. Post Hoc Analysis

       • Group Comparisons: If significant effects are found, conduct post hoc tests (e.g., Tukey’s HSD) to identify specific group differences.

       • Effect Sizes: Calculate effect sizes (e.g., Cohen’s d) for significant comparisons to understand the magnitude of differences.

       • Summary Table: Create a table summarizing means, standard deviations, p-values, and effect sizes for each group comparison.

    7. Visualization

       • Generate a graph displaying group means with 95% confidence intervals. Ensure the graph has clear labels for axes, group identifiers, and error bars. Remove any unnecessary background elements for clarity.

    8. Writing the Results Section

       • Formatting: Use APA or AMA style guidelines—double-spaced, Times New Roman 12-point font, with centered and bolded ‘Results’ heading.

       • Content Structure:

         • Briefly describe the study’s purpose and variables.

         • Summarize data screening processes and any issues addressed.

         • Detail the results of assumption tests.

         • Present ANOVA findings, including F-statistics and significance levels.

         • Discuss post hoc results and effect sizes.

         • Refer to the accompanying graph (e.g., “See Figure 1”) to illustrate findings.

    By following this structured approach, you’ll be able to systematically analyze the dataset and effectively communicate your findings regarding the impact of the Test Prep Course and socioeconomic status on third-grade students’ math performance in the STAAR assessment.

The post Analyzing the Impact of Test Prep Courses and Socioeconomic Status on STAAR Math Scores appeared first on Skilled Papers.