Instructions

You must follow the instructions below to get credits for this assignment.

Q1 Describe the Early Childhood Longitudinal Study that the U.S. Department of Education undertook in the late 1990s.

Hint: Make sure to discuss study’s goal, subjects, and variables in the data.

This was a study done on around 20,000 children between kindergarten and fifth grade. This study examined many different variables around the children such as their parent’s education levels, the parent’s income, whether they lived in a suburb, whether they had siblings, their test scores, and an assortment of other variables. This test was to find correlation or relationships between these variables.

Q2 Describe regression analysis in your own words.

Hint: A correct answer must have a discussion on a main concept of regression, often called as, “all else being equal”, “controlling for other variables”, or “Ceteris paribus”. The author explained this concept using “the circuit board analogy”.

Regression analysis is a way to find correlation or relationships between variables that are amongst a variety of other variables. It does this by evening the playinf field in all other variables except the ones being tested. For example, in the ECLS study one test they did was to see if african american children scored the same as white or asian children. The study gathered so many variables, that of course the correlation between these races were skewed. What it ultimately came down to was the fact that more african american children came from low income housing, and went to worse schools. Using regression analysis, the researchers were able to “level the playing field” by making all of the other variables constant. After doing this, they found that african american children scored just as well as white children if they went to similar schools.

Q3 What is a drawback of regression analysis. What type of questions can regression analysis not answer?

Hint: A correct answer must have a discussion on causality versus correlation.

Regression analysis finds correlational relationships between two variables, amongst a multitude of other variables. Therefore, regression analysis cannot find whether a relationship is causal. No outcome of regression can explain whether a variable caused another variable, although it can lead to that discovery depending on the relationship.

Q4 What role does the quality of schools play in academic performance of students.

Hint: See page 150.

In one study that used data from the ECLS study, researchers wanted to see if black children scored as highly as white children did. Upon first glance, they found that black children did not score as high as white children. After more depth in their research they discovered that there are a great quantity of black children who live in low income families, and therefore go to worse schools. This brought up the fact that because a large portion of black children go to worse schools compared to white children, there collective scores would be worse. This showed the importance of quality of schooling when comparing scores.

Q5 Continued from Q4. How could you control for the quality of schools in the study?

Hint: For this question, you may need additional information in addition to the assigned reading. You may Google search someting like “how does regression control for variables”.

Try to find other variables that are similar or the same around schooling before comparing the scores. For example, try to compare the scores of black and white children who have similar monetary class, or who’s parents make similar incomes. Find similar variables such as the number of parents who are present in the childs life. Variables of these nature to try and “level the playing field” before comparing the scores of the children.

Q6 The author says that regression is more art than science. What does he mean?

He means that it takes outside of the box thinking in order to connect the variables that matter. Regression is used when there are a multitude of variables, that can even seem random. That’s why it’s important to know how to piece together the variables and data in order to find true relationships. In a scientific study, it is relatively easy to create a control group and manipulate another variable to find the results of the manipulation. With so many variables involved with correlational studies, it is very difficult to create a control group. That is, until you can utilize regression analysis.

Q7 What are major takeaways from the reading?

Regression analysis is a very important tool if you want to find real and true correlational relationships between variables. Except in a perfect world, there will always be a multitude of variables that affect each other and it is very hard to compare any two variables because of these outlying ones. Using regression analysis, we are able to find the true relationship between two variables.

Q8 Hide the messages, but display the code and its results on the webpage.

Hint: Use message, echo and results in the chunk options. Refer to the RMarkdown Reference Guide.

Q9 Display the title and your name correctly at the top of the webpage.

Q10 Use the correct slug.