EC315 – Quantitative Research Methods

Fall II 2011

TABLE OF CONTENTS

BACKGROUND…………………………………………………………………………………………3 REGRESSION ANALYSIS……………………………………………………………………..………4 CONCLUSIONS…………………………………………………………………………………………6 BIBLIOGRAPHY………………………………………………………………………………………..8 APPENDIX……………………………………………………………………………………………….9

I. Background

According to the Center for Disease Control and Prevention (CDC)” Type 2 diabetes, which was previously called non-insulin-dependent diabetes mellitus (NIDDM) or adult-onset diabetes, may account for about 90% to 95% of all diagnosed cases of diabetes”. The CDC also attributes the onset of type 2 diabetes to obesity in many cases. The purpose of this analysis is to determine the effects of obesity (OBESE) on the incidence of diabetes (DIABETIC) while holding the effects of alcohol consumption (ALCOHOL), ethnicity (HISPANIC), and age (AGE) constant. This study will use cross-sectional data from the 50 states for the year 2010. The model (less constant and coefficients) is: DIABETIC = OBESE + ALCOHOL + HISPANIC + AGE

The dependent variable, DIABETIC, is defined as the percent of people who have been told by a doctor that they have diabetes, and is extracted from the Center for Disease Control and Prevention Behavioral Risk Factor Surveillance System (2011).

Data for the independent variables OBESE, ALCOHOL, HISPANIC, and AGE were also taken from the Center for Disease Control and Prevention Behavioral Risk Factor Surveillance System (2011) by percentage. OBESE is defined as individuals with a body mass index (BMI) greater than 30. ALCOHOL is defined as heavy drinkers (adult men having more than two drinks per day and adult women having more than one drink per day). HISPANIC is defined as individuals who identify themselves as having Hispanic ethnicity. Lastly AGE is defined as individuals 45 years of age and older. The independent variables were selected due to their association with diabetes. According to the CDC “Risk factors for type 2 diabetes include older age, obesity,… and race/ethnicity” (Center for Disease Control and Prevention, 2011). The American Diabetes Association links alcohol consumption and hypoglycemia “Women should drink 1 or fewer alcoholic beverages a day (1 alcoholic drink equals a 12 oz beer, 5 oz glass of wine, or 1 ½ oz distilled spirits (vodka, whiskey, gin, etc.). Men should drink 2 or fewer alcoholic drinks a day” (American Diabetes Association, 2011).

II. Regression Analysis

The model was regressed and the results are shown in Table 1. Table 1 - Original Regression Results

Dependent Variable: DIABETIC Adjusted R2 = 0.6442 n = 54 | Independent Variables| Coefficients| t Statistic| P-Value| OBESE| 0.3486| 6.4528| 0.0000|

ALCOHOL| -0.2904| -1.8262| 0.0739|

HISPANIC| 0.0441| 4.1307| 0.0001|

AGE| 0.1207| 4.3074| 0.0001|

The initial regression analysis is pretty good. The adjusted R2 value of 0.6442 shows there is a moderate positive correlation. The coefficients for OBESE, HISPANIC and AGE are positive as expected. The coefficient for ALCOHOL, however, is negative where a positive result was expected. This might be because while ALCOHOL has negative effects on a person suffering from diabetes it is not itself a cause of diabetes. Using a level of significance of 0.05 and using the P-Value above all variables with the exception of ALCOHOL are shown to be statistically significant.

Following the initial regression a correlation matrix was completed. The results are shown in Table 2. Table 2 - Cross Correlation Matrix

The data above shows a good cross correlation matrix. There are no coefficients less the -0.7 or greater than 0.7. This indicates that no multicollinearity was found using the cross correlation matrix.

Variance inflation factors were calculated. The results are shown in Table 3.

Table 3 - Variance Inflation Factors

All variance inflation...