Description

Session Long Project

The purpose of this paper is to examine predictors for the number of weekly hours spent using email through analyzing descriptive statistics, correlation and bivariate analysis, checking assumptions, and performing linear and multiple regression.

Methodology

There are six variables in the data set. First, ID represents the ID number of a respondent. Second, AGE denotes the age of a respondent. Third, CHILDREN represent the number of children a respondent has. Four, SEX represents the sex of a responder. Five, EMAILHR signifies how many hours per week a respondent uses their email. Six, INCOME embodies the total family income. Out of these six variables, EMAILHR is the dependent variable, and the others are independent variables. Table 1 summarizes the variables in this study, providing the name of the variable, whether the variable is a dependent one or independent, and the type of the variable, such as categorical, ordinal, or continuous.

Table 1

Variables Used in the Study

Descriptive Statistics

The age variable ranges from 20 to 76 years old. The mean age of the respondents is 45.13 years old, whereas the median age is 43.00 years old. This tendency indicates that the distribution of data is skewed to the right because it is typical for right-skewed data to have the median smaller than the mean. In other words, the sample tends to have slightly more younger respondents than older ones. The standard deviation for the AGE variable is 14.571, which implies there is a considerable variation among the respondents in their age.

The number of children ranges from 0 to 7. The mean the number of children is 1.69 children per respondent, whereas the median number of children is 2.00 children. Although the mean is less than the median, the data is undoubtedly right skewed. This conclusion is because many individuals do not have children at all, and in the modern world, families rarely have more than two children. The standard deviation, SD = 1.638, also proves that there is a wide variety among the respondents.

The sample (n = 87) has 46 women and 41 men. Consequently, females make up 52.87 percent of the sample, and males make up 47.13 percent. As for the level of income, most respondents (71.26 percent) earn $25,000 or more per year, and only 28.74 percent make between $1,000 – $24,999.

The weekly hours spent using email range from 0 hours to 60 hours. The mean number of hours is 8.78 hours per week, whereas the median is only 2.00 hours weekly. This enormous difference between the mean and median suggests that the data is skewed to the right. Indeed, the most common value in the EMAILHR variable is 0, which means many individuals do not use emails at all. The standard deviation is 13.885, which is even higher than the mean.