Big Business Support
Introduction
Among many different things in our lives today between jobs and others areas of life people in the United States are affected by every day big business is one area that I am curious of. With the financial standpoint of Americans I hold a curiosity as to what affects people opinions of it and what factors may be involved. Big business is involved in everything, how we get our food to how we interact with people this is all influenced by those large corporations that say what is favored in society. As Political Science students we have interests in politics which is largely controlled by those with wealth or those that hold high positions in larger corporations or in big businesses. This is an important factor which is why it is so interesting to me. Based on this I am hoping to find out what people’s opinions are on big business and whether or not their education level, employment status and political ideology affects these opinions. For all of the following the dependant variable used to compact each variable against is the variable bigbus_therm which is an interval variable from the NES 2004 data set which all of the following variables were used from. This variable is a feelings thermometer which ranks the opinions on a ranked scale of the lowest number being completely against to the highest number being in complete support.
The first of these factors is the education level of those who were surveyed. People who have a higher level of education are more likely to support big business than those who have a lesser level of education. I have come to this conclusion because with more education are more likely to be involved with larger businesses and they are able to see how important these big businesses are to our economy and how without them it is detrimental to society. The variable am using is educ7 from the NES 2004 data set and is divided into 7 different category’s. This category’s range from little high school all the way to multiple or a single master’s degree from a university in any area. For these categories, involving all of them is important because as one becomes more educated their opinions change from high school to an undergraduate degree or even a masters/graduate degree. There is also partial education included because there may have been underlying conditions involving their education level that may affect the outcome of these test and because this variable is ordinal from the levels of education categorized as an increasing of their education the category’s show the change of opinion as the education increases.
The second factor I am researching is the employment status of those who were surveyed. The variable used is called employstat from the NES 2004 data set as well. For this data set it was somewhat confusing in some areas of the variable because of the fact that there are several different divisions that are specific to unknown reasons. Some of these factors where disabilities, injury and those who may have been laid off for different reasons, For the difference of means test I only wanted to measure the opinions of those who were either working or unemployed because those in other category’s of the variable may have biased opinions based on their current status. For the rest of the data I used the entire data set to show the differences in these opinions and to represent all areas equally. But in for each test my hypothesis stayed the same; people who are employed are more likely to support big business than those are unemployed. This data set is ordinal based that there is no way to clearly categorize this data including all of the special situations.
The last variable used is for the political ideology of the individual named libcon3_r from the NES 2004 data set. A way to describe this is that; people who are Republican are more likely to support big business than those who are not Republican. This variable is coded into 3 separate areas being; Republican, Democrat, and independent, of which could be seen as Republican more likely supporting big business, where as independents usually support big business less and Democrat believe more in equality than either of the other two. The political ideology of the person can affect their opinions because of their opinions on different policies that affect big business or the public. Such as for example environmental restrictions, each ideology may have different stances on the subject and therefore the impact it has on big business would be different for each ideology. The variable is coded clearly as an ordinal variable because there is no distinct order that could be formed and because of that the opinions will be clearly distributed.
Other Variables
Of these 3 independent variables I have chosen to compare to the big business dependant variable there are many other factors that may affect ones opinion of big business. Two of these variables may be the person’s age or their media exposure. When considering a person’s age it could be assumed that; people who are older than 65 are less likely to support big business than those who are under 65. This may be because those who are 65 and older may have already retired and because of that they may not support big business controlling large amounts of a specific industry as well they have had more experience with big business and their age may reflect the experiences they have had with it. For the other variable; people who are exposed to more media are more likely to support big business than those who are exposed to less. This could be because the media is able to control public opinion based on what is released so there would be a more positive view reflected by different forms of media because of big businesses ability to influence it. These could also be affect different people’s opinions on big business but there are too many underlying variable that can affect the specific affect of these variables in which are too complicated to test such as who has had positive life influences with big business for their age or whether they are involved in big business and the media may not be relevant to their area of big business.
Data
When starting my research I conducted a difference of means test using the Independent Sample T test which compares the difference of the means of the selected independent variable’s category’s to represent the difference in the average score or select in the specified variable also measuring whether or not the hypothesis can be supported or if there is too high a level of error involved in the test to differ. By conducting the difference of means test I used the variable employstat as my independent variable and recoded it into two different categories’ using 1 as working now and 4 as currently unemployed. The dependant variable used was bigbus_therm. For this specific test I wished to test whether people who are currently employed are more likely to support big business than those who are currently unemployed. According to the data there is only a difference in the means of 1.030. Using the Sig value of 0.005 I am able to use the bottom row of the variances assumed for my Sig (2-tailed0 value). But based on the Sig (2-tailed) column my conclusion is unsupported. According to the Sig (2-tailed) there is a value of 0.803 which means there is just over an 80% chance that the results were concluded due to random error. This means that my hypothesis is rejected and that the null hypothesis is accepted because of the fact that there is no relationship between whether the employment status of a person affects their opinion on big business.
Difference of Means T-test employ status 1-employed 4-unemployed
Group Statistics
R employment status N Mean Std. Deviation Std. Error Mean
Feeling Thermometer: Big Business dimension1 1 Working now 695 55.17 21.346 .810
4 Unemployed 29 54.14 30.062 5.582
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 95% Confidence Interval of the Difference
Lower Upper
Feeling Thermometer: Big Business Equal variances assumed 8.004 .005 .250 722 .803 1.030 4.122 -7.063 9.123
Equal variances not assumed .183 29.190 .856 1.030 5.641 -10.503 12.564
The next level of testing is that I conducted a bivariate regression between each of my independent variables and my dependent variable. The first variable I tested for is educ7 which is education level of those surveyed. The first area to be looked at is R value in the model summary. This value is measured at 0.08 which means that only 8% of the variance in our DV can be explained by the education level of the person. This means that the relationship between the IV and the DV is not were significant. This means that my hypothesis is most likely incorrect because of the little relationship there is between ones education and their opinions on big business. Then when looking at the coefficients table the Sig column has a 0.010. Being as the number is less than 0.05 the null hypothesis is rejected so that this data is caused by random error. When looking at the B column or beta value the number is -1.078. This means that for every increase of the independent variable by 1 the dependent variable’s score decreases by -1.078 for the little effect it has on it. This shows there is a negative relationship between the education level of the people surveyed and their opinions on big business.
Bivariate Regression Educ7
Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
dimension0 1 .080a .006 .005 21.505
a. Predictors: (Constant), R education level: 7 cats
Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig.
B Std. Error Beta
1 (Constant) 60.334 1.921 31.410 .000
R education level: 7 cats -1.078 .416 -.080 -2.592 .010
a. Dependent Variable: Feeling Thermometer: Big Business
For the variable employstat I conducted another bivariate regression to measure the relationship between the employment status of those surveyed and their opinions on big business. For this test I again looked at the R value in the model summary and it was measured as 0.036. This means that 0.36% of the variance can be attributed to the employment status of the person. This is a weak relationship meaning ones opinion on big business is not affected by their employment status. After learning of this you look at the Sig value of the coefficients table to learn that the Sig value is 0.240. Although because this value is greater than 0.05 the null hypothesis fails to be rejected.
Bivariate Regression Employstat
Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
dimension0 1 .036a .001 .000 21.570
a. Predictors: (Constant), R employment status
Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig.
B Std. Error Beta
1 (Constant) 54.787 .998 54.905 .000
R employment status .346 .294 .036 1.177 .240
The last variable I ran a bivariate regression was the libcon3_r which measures the political ideology of the peoples surveyed. For this variable I chose to leave the variable as is so that I would be able to notice if there was a change between any of the three ideologies. From the model summary in the R column I am able to deduce that from the 0.326 value there is a strong relationship and that just over 32% of the variance can be contributed to the ideology of the person surveyed. After this I looked at the Sig value in the coefficients table to a 0.00 value meaning because it is below 0.05 the null hypothesis is rejected. This mainly means that random error did not cause the results. So when looking at the B column you can see that the beta value is 8.797. This means that for each level of increase by 1 in ones political ideologies the persons support for big business increases by 8.797 points, so approximately 9 points. Based on this for the level of 3 which is the max and also the republican value it means that one’s support increases the most for being Republican versus Democrat or independent. This means my hypothesis is supported.
Bivariate Regression libcon3_r
Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
dimension0 1 .326a .106 .105 20.691
a. Predictors: (Constant), Ideology: 3 cats
Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig.
B Std. Error Beta
1 (Constant) 36.362 2.076 17.513 .000
Ideology: 3 cats 8.797 .897 .326 9.812 .000
a. Dependent Variable: Feeling Thermometer: Big Business
For my final test I conducted a multiple regression analysis to see if the independent variables I chose affect each other in any other way to increase the percent of variance or their beta value as to how much the relationship affected an increase or decrease in support. Based on the model summary the R value is 0.332 meaning that these 3 variables affect about 33% of the dependent variables variance. As I see it this is not a major difference because of each of my previous bivariate regression analyses the political ideology affect 32% of the variance so there is no major difference between these variables based on the R value and the strong relationship is based on the variable libcon3_r. For the coefficient table the Sig value for education level and employment status is still above 0.05 so it can be accepted that both are caused due to random error again and the null hypothesis is accepted meaning each has no effect in ones opinions on big business while the political ideology variable rejects the null hypothesis.
Multivariate Regression
Model Summary
Model R R Square Adjusted R Square Std. Error of the Estimate
dimension0 1 .332a .110 .107 20.685
a. Predictors: (Constant), Ideology: 3 cats, R employment status, R education level: 7 cats
Coefficientsa
Model Unstandardized Coefficients Standardized Coefficients t Sig.
B Std. Error Beta
1 (Constant) 39.754 3.166 12.556 .000
R education level: 7 cats -.795 .462 -.057 -1.721 .086
R employment status .159 .330 .016 .482 .630
Ideology: 3 cats 8.730 .897 .323 9.731 .000
a. Dependent Variable: Feeling Thermometer: Big Business
Conclusion
In conclusion, based on the weak percent variance and the small beta value (B) and Sig value for employment status and education level for the regression analysis, I conclude that for both variables the null hypothesis is accepted and both hypotheses are incorrect due to random error and the weakness of the relationship. On the other hand the data for the political ideology regression analysis supports my hypothesis through the fact that the positive relationship between those who are republican and those who are of a different ideology has a greater affect on the support of big business. Also the fact that the Sig is 0.00 means that the null hypothesis is rejected unlike my other two independent variables. After completing the multiple regression analysis both the education level and employment status of those surveyed are both rejected as being due to random error while the only hypothesis still remaining supported is of political ideology. From the data collected those who are Republican are more likely to support big business than those who are democrat or independent.