All papers examples
Get a Free E-Book!
Log in
HIRE A WRITER!
Paper Types
Disciplines
Get a Free E-Book! ($50 Value)

Linear Regression, Essay Example

Pages: 5

Words: 1245

Essay

Independent T-Test Samples

Independent Samples Test
Levene’s Test for Equality of Variances t-test for Equality of Means
F Sig. t df Sig. (2-tailed) Mean Difference Std. Error Difference
CTBS Language NCE score Equal variances assumed .597 .440 3.016 314 .003 6.01259 1.99351
Equal variances not assumed 3.012 310.703 .003 6.01259 1.99616

One-Way Anova

ANOVA
CTBS Language NCE score
Sum of Squares df Mean Square F Sig.
Between Groups 2854.121 1 2854.121 9.097 .003
Within Groups 98517.691 314 313.751
Total 101371.812 315

General Linear Regression

Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
1 .168a .028 .025 17.71301
a. Predictors: (Constant), gender

 

Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig.
B Std. Error Beta
1 (Constant) 52.994 1.392 38.080 .000
gender -6.013 1.994 -.168 -3.016 .003
a. Dependent Variable: CTBS Language NCE score

The t-test is simply comparing the language score means between males and females based on the correlational coefficient. In this example, the female mean score (52.990) is higher than the male mean score (46.98), when the size of the sample is taken into account, a t-score of 3.01 is calculated (p-value>.001).

One-way ANOVA is also a correlational analysis, based on the correlational coefficient between variables. The ANOVA analysis comes up with the exact same p-value as the t-test (.003); however, it uses an f-test rather than t-test.

General linear regression gives substantially more information than a t-test or ANOVA. The parameter “gender” is significant ( a face one would know from the previous two tests); however, the regression allows to know the correlation between the two (negative) and the approximate effect of gender on tests.

Univariate Analysis

Tests of Between-Subjects Effects
Dependent Variable:CTBS Language NCE score
Source Type III Sum of Squares df Mean Square F Sig.
Corrected Model 11678.496a 9 1297.611 4.427 .000
Intercept 122302.969 1 122302.969 417.252 .000
gender 42.755 1 42.755 .146 .703
engprof 8410.914 4 2102.728 7.174 .000
gender * engprof 1263.217 4 315.804 1.077 .368
Error 89693.316 306 293.115
Total 893388.982 316
Corrected Total 101371.812 315
Multiple Regression

Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
1 .317a .100 .092 17.09616
a. Predictors: (Constant), gen_prof, gender, profiiency in native language

 

Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig.
B Std. Error Beta
1 (Constant) 56.851 1.632 34.833 .000
gender -6.939 2.381 -.194 -2.915 .004
profiiency in native language -5.433 1.306 -.334 -4.160 .000
gen_prof (interaction) 2.139 1.762 .111 1.214 .226
a. Dependent Variable: CTBS Language NCE score

The main difference between the two models is that the unvariate analysis gives on whether the factor (gender or bilingual) was a significant predictor of the language score (via the F-score). The linear regression model actually gives parameter estimates for the main factors (gender, proficiency. Due to the presence of categorical variables, this interpretation is not intuitive. Gender (-.194) is actually the parameter for being male (“male=1”) meaning that being male lowers one test scores. Similarly, there is a negative interpretation of proficiency in native language variable. Finally, the interaction variable is significant.

The average mark in algebra (dependent variable) is skewed, largely because of the decision to group the variables by grades rather than keeping the continuous scores- a large number of students that failed (112) and earned a d, but the decision to model the variable in this way essentially puts high fails and low “d”s in the same category. The CTBS Math NCE score looks to be roughly normally distributed- based on the decision to keep it as a continuous score. The “school” variable is even distributed between school alpha and school beta. Finally, number of days absent is positively skewed: this is a function of numerous students not missing any days of school (0), and the distribution thinning as the number of absent days increases. Although there are some problems with this data, the regression is generally robust.

Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig.
B Std. Error Beta
1 (Constant) .728 .236 3.079 .002
number of days absent -.047 .009 -.271 -5.376 .000
CTBS Math percentile .022 .003 .439 8.592 .000
mark in spring math .137 .051 .135 2.670 .008
school -.310 .134 -.121 -2.307 .022
a. Dependent Variable: mark in algebra

Looking at the results, all of the parameters are significant, and most of the results make intuitive sense: an increase in the number of days absent leads to a lower algebra score; grades in the CTBS and mark in spring math are positively correlated with the score in algebra; finally, depending on the school chosen, there is a negative correlation. Looking at the effect size, the mark in spring math is a strong positive predictor; and school a strong negative predictor.

The difference between unstandardized coefficients and standardized coefficients is the use of scale: standardized coefficients are corrected for scale. The main difference between the two are the strong negative predictor (number of days absent) and the strong positive predictor (CTBS Math Percentile). Finally, school is less important once scaled to the other variables

I chose to add on the model above to understand how variables that look at English grades and bilingual status might also predict math grades. The intuition behind this inclusion is that individuals that don’t understand English may not understand what the math instructor is saying; or may also not ask needed questions due to a perceived obstacle in communication with the teacher.

Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
1 .660a .436 .425 .973
a. Predictors: (Constant), number of days absent, profiiency in native language, CTBS Math NCE score, mark in English 95, bilingual status, CTBS Language NCE score

Model= Algebra Scores (DV)= Mark in English (IV)+ CTBS Math Score (IV)+ CTBS Language NCE Score (IV) + Bilingual status (IV)+ Proficiency in native language (IV)+ Number of days absent

Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig.
B Std. Error Beta
1 (Constant) -.504 .233 -2.165 .031
mark in English 95 .458 .052 .448 8.793 .000
CTBS Math NCE score .027 .004 .377 6.367 .000
CTBS Language NCE score -.008 .004 -.108 -1.738 .083
bilingual status -.024 .060 -.021 -.407 .684
profiiency in native language .126 .059 .108 2.138 .033
number of days absent -.021 .008 -.124 -2.623 .009
a. Dependent Variable: mark in algebra

Overall, this model has a significantly higher adjusted r-squared than the previous model- this is important because the regular r-squared will increase regardless of whether the variable is significant or not. My hypothesis that language plays a role in math scores had a mixed result: “Mark in English 95” was a significant positive predictor of the algebra score; this could be because it is capturing that good students do well, or it could be that those with better English ability can understand math better. “Proficiency in native language” was also a significant positive predictor leading one to believe that language ability plays some type of role in understanding math. The other two variables, “CTBS language score” and “Bilingual Status” were not significant in predicting the algebra score.

Time is precious

Time is precious

don’t waste it!

Get instant essay
writing help!
Get instant essay writing help!
Plagiarism-free guarantee

Plagiarism-free
guarantee

Privacy guarantee

Privacy
guarantee

Secure checkout

Secure
checkout

Money back guarantee

Money back
guarantee

Related Essay Samples & Examples

Relatives, Essay Example

People have been bound by bloodline and kinship since times immemorial. This type of relation is much more complex than being simply unified by common [...]

Pages: 1

Words: 364

Essay

Voting as a Civic Responsibility, Essay Example

Voting is a process whereby individuals, such as an electorate or gathering, come together to make a choice or convey an opinion, typically after debates, [...]

Pages: 1

Words: 287

Essay

Utilitarianism and Its Applications, Essay Example

Maxim: Whenever I choose between two options, regardless of the consequences, I always choose the option that gives me the most pleasure. Universal Law: Whenever [...]

Pages: 1

Words: 356

Essay

The Age-Related Changes of the Older Person, Essay Example

Compare and contrast the age-related changes of the older person you interviewed and assessed with those identified in this week’s reading assignment. John’s age-related changes [...]

Pages: 2

Words: 448

Essay

The Problems ESOL Teachers Face, Essay Example

Overview The current learning and teaching era stresses globalization; thus, elementary educators must adopt and incorporate multiculturalism and diversity in their learning plans. It is [...]

Pages: 8

Words: 2293

Essay

Should English Be the Primary Language? Essay Example

Research Question: Should English be the Primary Language of Instruction in Schools Worldwide? Work Thesis: English should be adopted as the primary language of instruction [...]

Pages: 4

Words: 999

Essay

Relatives, Essay Example

People have been bound by bloodline and kinship since times immemorial. This type of relation is much more complex than being simply unified by common [...]

Pages: 1

Words: 364

Essay

Voting as a Civic Responsibility, Essay Example

Voting is a process whereby individuals, such as an electorate or gathering, come together to make a choice or convey an opinion, typically after debates, [...]

Pages: 1

Words: 287

Essay

Utilitarianism and Its Applications, Essay Example

Maxim: Whenever I choose between two options, regardless of the consequences, I always choose the option that gives me the most pleasure. Universal Law: Whenever [...]

Pages: 1

Words: 356

Essay

The Age-Related Changes of the Older Person, Essay Example

Compare and contrast the age-related changes of the older person you interviewed and assessed with those identified in this week’s reading assignment. John’s age-related changes [...]

Pages: 2

Words: 448

Essay

The Problems ESOL Teachers Face, Essay Example

Overview The current learning and teaching era stresses globalization; thus, elementary educators must adopt and incorporate multiculturalism and diversity in their learning plans. It is [...]

Pages: 8

Words: 2293

Essay

Should English Be the Primary Language? Essay Example

Research Question: Should English be the Primary Language of Instruction in Schools Worldwide? Work Thesis: English should be adopted as the primary language of instruction [...]

Pages: 4

Words: 999

Essay