Descriptive Statistics

 

How do students with varying levels of English proficiency interact with computerized math assignments in English?

 

Research Questions


I used this set of data from my own classroom students. I have created the scenario and analyzed the data below. The subjects of the study are students from my own classroom. The data was collected from student activities between July 23, 2003 and March 11, 2004.

With my classroom's implementation of a computerized math assignment program that is only available in English, I wanted to see how my students would perform on this new program.   Specifically, I wished to examine the variety of English proficiency levels in my classroom, how my Mainstream English Only (EO) students performed on the program, and how my English Language Learners (ELL) performed on the program.   In order to better understand this classroom situation, I looked for significant correlations between these variables.

Data Collection and Analysis


Data were collected through school registration records, the school nurse health office, classroom computerized assessments, and current California State English Assessments. School registration records detailed the gender of the students, their dates of birth, their home language, the date each student entered the school, and their lunch program status.   The school nurse health office provided information regarding students use of glasses, any students taking daily medication, and any students that suffered from asthma.   Classroom assessments included each student's language arts instructional program, their language of instruction in language arts, the number of Math Test Check assignments assigned in English, the number of Math Test Check assignments completed in English, and the average percent correct on a Test Check assignment in English.   Current California State English Assessments provided overall scores and proficiencies as well as scores and proficiencies in the categories of listening and speaking, reading, and writing for the students who took the California English Language Development Test (CELDT) in October of 2003. I chose the Reading Section since that is the skill that the students use while working on the computerized math assignments.   Also, there is no computerized vocalization of the math problems while students work on the assignments so students are not able to use their listening and speaking skills to assist them in solving the problems on the assignments.   These data were entered into Excel and SPSS 11 for descriptive analysis.

Descriptive Analysis


The creators of the CELDT score have established the below ranges of proficiency for the students that take the test:

Student Score

Proficiency Level

0 to 449

Beginning

450 to 499

Early Intermediate

500 to 549

Intermediate

550 to 599

Early Advanced

600 and above

Advanced

 

Calculating the mean for the 23 ELL students that took Reading Section of the CELDT shows that the 23 ELL students have a mean score of 470.1 (SD=44.42).   The range of the scores is 191 and the distribution appears to be evenly distributed.

The mean of 470.1 suggests that the proficiency level of the 23 ELL students is "Early Intermediate."     Next, I look at the performance of the 23 ELL students on the computerized math assignments.   I see that the mean for the 23 ELL students is 57.1 (SD=16.16).  

Correlational Analysis


I hypothesized that the students with the higher CELDT reading score would have a higher percentage of correct answers on the computerized math assignments.

If I look at the scatterplot, I see a positive correlation between the students CELDT reading scores and the average percent correct on the computerized math assignments.

Correlational analysis using the Pearson R, of CELDT Reading Score to Average Percent Correct on the Computerized Math Assignments, reveals a positive correlation between the two variables (r=0.537).    

Conclusion


The students with higher English proficiency performed better on the computerized math tests that were only available in English.   The correlation between the student CELDT reading scores and their average percent correct on the computerized math assignments may be do to the ability to read the math questions in English on the computer.