Introduction

We will be using the data from the 1985 Current Population Survey which was complied of random samples from people, and we will be referring to this data known as CPS85 which has 245 females and 289 males. Altogether the survey consisted of 534 observations with these variables:

We will be analyzing the variable sex, among other confounding factors because it’s believed that there’s discrimination in the work place between genders.

Methods

Now we will decide whether the variables are numerical or a factor so that we may make suitable charts for this data.

Variables:

This is just a further breakdown of the variables that we will be examining based on how they will be prepared in the results. They were identified as explanatory or response to see which is the contributing factor.

Relationship Variables:

Results

Wage and Sex

Here there’s a greater density observed on the females side, and on the males’ side there’s a much larger cluster for bigger amounts of dollars per hour.

##   sex  min   Q1 median Q3   max     mean       sd   n missing
## 1   F 1.75 4.75   6.80 10 44.50 7.878857 4.720113 245       0
## 2   M 1.00 6.00   8.93 13 26.29 9.994913 5.285854 289       0

This is the data we want to analyze and decide if it’s discrimination between gender or if there’s a confounding variable because there is a notable difference for males making more money than females.

Sex and Education

We’ll be looking at data to answer our question from earlier: Is education a confounding variable in the relationship between sex and wage?

##   sex min Q1 median Q3 max     mean       sd   n missing
## 1   F   6 12     12 15  18 13.02449 2.429205 245       0
## 2   M   2 12     12 15  18 13.01384 2.767586 289       0

Here we can see that men and women are fairly equal with years of education, and there’s even more men on the left side of the male data than the female’s, meaning they have more with a lower amount of education. Of course, the more education you have, the more wage you’ll make. It can be concluded that education isn’t a confounding variable since both genders are quite equal.

Sex and Sector

Now we will analyze the data to decide what the answer is to our question: Does sector of work play as a confounding variable in the relationship between sex and wage?

##    sector
## sex clerical const manag manuf other prof sales service
##   F       76     0    21    24     6   52    17      49
##   M       21    20    34    44    62   53    21      34

Notice how there’s no females in construction.

Here are the wages of each sector.

##     sector  min     Q1 median      Q3   max      mean       sd   n missing
## 1 clerical 3.00 5.2000  7.500  9.5000 15.03  7.422577 2.699018  97       0
## 2    const 3.75 7.2250  9.750 11.6275 15.00  9.502000 3.343877  20       0
## 3    manag 1.00 7.1250 10.620 15.8550 26.29 12.115185 6.244713  54       0
## 4    manuf 3.00 4.9250  6.750  9.8725 22.20  8.036029 4.117607  68       0
## 5    other 2.85 5.0000  6.940 10.8150 26.00  8.500588 4.601049  68       0
## 6     prof 4.35 7.5000 10.610 15.3800 24.98 11.947429 5.523833 105       0
## 7    sales 3.35 4.3125  5.725 10.8325 19.98  7.592632 4.232272  38       0
## 8  service 1.75 3.9650  5.500  8.0000 25.00  6.537470 3.673278  83       0

Even when we compare wage and sector along with gender, we still see that males make more than females. There are more males in the survey so their numbers are much larger on each sector except for clerical and service. There’s 245 females in the survey and 289 males in the survey. Males tend to be in construction and manufacturing and women are more in the service and less masculine jobs. Sector of work can be a confounding variable because this evidence shows this.

Wage and Race

Here we will analyze wage and race to decide the answer to our question: Would race be a confounding variable in the relationship between sex and wage?

It’s clear that whites make much more than nonwhites.

Whites have a larger spread than nonwhites, with even more outliers as well.

##   race  min    Q1 median     Q3   max     mean       sd   n missing
## 1   NW 3.35 4.710   7.50 10.000 23.25 8.058358 4.078379  67       0
## 2    W 1.00 5.255   7.81 11.395 44.50 9.162612 5.262928 467       0

It’s clear that from this data that there is a relationship between race and wage. Whites tend to make more money than nonwhites, and this could be the confounding variable in our data.

Discussion

When we looked at sex and education, we saw that both males and females had about the same amount of years of education. The males did have more people starting off with lower years of education than females, which questions why it is that males are still making a noticeable amount of money, which is more than females. Other than that, education isn’t really a confounding variable because both genders are quite equal.

Sex and sector revealed that there were absolutely no women in construction, and every sector had more males than women except for clerical and service. Even the wages were different for the genders when they were in the same sector of work! This would’ve been a confounding variable except that the men were getting paid more than women in the same job.

There was quite a difference in the wage and race data. Whites were making much more than nonwhites. This is a confounding variable because it shows that race affects wage.

The data could’ve had factors not accounted for.

We have concluded that there is discrimination in work wages between sex. Males make more than women despite having the same amount of education, and after observing the sector wages for both genders, there was evidence that men made more than women.

Something that we would’ve liked to research more in depth but didn’t have time to do was analyze the relationship between race against wage and sex. Race played a part in the wage difference, so is it also playing a part in this data as well?