Relationship between Attitudes toward Abortion and Number of Children
Introduction to Statistics for Sociology and Social Work
Marina Phal
Pacific Lutheran University

Relationship between Attitudes toward Abortion and Number of Children
This research looks at the relationship between how one feels about abortion and the number of children they have. The research aims to assess whether there is any causality connection between the two variables. The preliminary conclusion of the research is that if one feels negatively towards abortion, the number of children they have would be higher than those who are for abortion and vice versa. This research can be a useful resource to better understand how families in United States decide on the number of children they should have in households, as well as to give an idea of whether or not it is the subjective feeling about abortion that affects how many children they should have.
METHOD
To process the research, the General Social Survey from 2012 is used. The GSS 2012 is a collection of survey data on demographic characteristics and attitudes of people living in United States. The data is comprised of a wide array of sociological data ranging from personal information, including sex, age, income and the like, to attitudes toward different issues, including abortion, racism, and so on. To achieve the purpose of this research, the dependent variable used is the actual number of children. This variable is measured by number ranging from zero to seven. Respondents who have eight children or more will be considered as only one category. The independent variable used in the research is attitudes toward abortion (abortion if women wants for any reason, as explained in the GSS). The responses to this variable are categorized as yes and no. The yes category is used as the omitted category. To ensure the relative impact of the independent variable, seven control variables will be held constant – highest degree earned, family income, ideal number of children, age, sex, subjective judgment on how religious the respondent think he/she is, and how often the respondent attend religious services. Highest degree earned is explained by the highest level of education attained. The original variable responses range from little high school to graduate. The variable will be recoded into only three categories including high school, undergraduate and graduate. Little high school and high school categories will be merged into one as well as junior college and bachelor. The reference category will be undergraduate. Secondly, family income is identified as the actual income the respondents earn in constant dollar. The originality of this variable will be kept. Similarly, the ideal number of children will be reserved the way it is except excluding “as many as want” category from the analysis to avoid influences on the outcome. This variable is also identified with the actual number of children varying from zero to six except considering number of children higher than or equal to seven as one category.
For age, actual number of age is presented in the data. Respondents whose ages are higher than 89 years old is counted as one category. Next, sex is categorized as either male or female. Male will be used as the reference category for sex. The sixth control variable is the subjective identification of how religious the respondent think he/she is. The original categories of this variable include very religious, moderate religious, slight religious, and not religious. These categories will be recoded into three only – very religious, religious (merged from moderate religious and slight religious), and not religious. The religious category will be considered as the reference category. The last control variable is more objective assessing how often the respondent goes to religious services. Originally, eight categories are included, ranging from little once a year to more than once a week. They will