You must follow the instructions below to get credits for this assignment.
Hint: Make sure to discuss study’s goal, subjects, and variables in the data.
The ECLS was a program where more than 20,000 young children between the grades of kindergarten and fifth grade, were asked simple demographic questions with a goal of finding their education level. Students were asked race, gender, family structure, socioeconomic status, parents level of educaiton, and more. But more information was gathered from teachers, principals, and parents about how often they were spanked if at all, how often they watch tv, are the children ever brought to educational recreation sites, and so on. In this study, there are hundreds of variables to be accounted for where in normal studies there are only a couple variables. Hundreds of variables would make it difficult to analyze in a normal setting but using regression analysis tools have made it much easier to read and comprehend.
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 kind of like grouping variables to see how and if they correlate. When using an experiment setting, usually all variables are controlled with a random subject(s) being altered. The regression analysis format can group subjects with likewise characteristics or with characteristics that they dont have in common. Regression analysis doesnt just answer any question though, it answer if X can be correlated to Y and vise versa. This still doesn’t prove that the question is 100% true, it basically says that there is reason to believe that X is correlated with Y because the numerics say so.
Hint: A correct answer must have a discussion on causality versus correlation.
Regression analysis can not answer questions that are considered casuality questions such as the one used in the text “does having a lot of books in your home lead your child to do well in school?”. This is basically a yes or no question that can’t be answered with regresssion analysis. Regression analysis is meant for answering questions that ask if variable 1 is related to variable 2 such as “does a child with a lot of books in his home tend to do better than a child with no books?”. Now using data, this question can be answered in terms of correlation where a child with books may do better than a child with no books. But there can be hundreds of other variables as to why the child with books does better in school such as family structure and socioeconomic status.
Hint: See page 150.
The quality of school plays a significant role in academic performance. In the survey, the typical white student attended schools where there was only a 6% black population and came from better neighborhoods and a much lower rate of “troublesome indicators”. The typical black student that was surveyed came from schools that were 60% black population but had a significant higher rate of troublesome indicators like gang relations, non-student loitering, and not much PTA funding. Regardless of technology, teacher education, or money, the two populations were very similar other than the troublesome indicators. This shows that the area that a student is in plays a significant role in how well they will perform academically.
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”.
Well in this case you cannot control the quality of schools for this study as this is an observational study. There is no altering to the students, rather they are just being examined for their academic ability. Observational studies do not have a control and usually have confounds. A confounding varibale is a variable that cold affect the X and Y but is not necassarily measured.
The thing that stuck out the most to me was the fact that you can do a regression analysis on any correlating question that you can think of. It is crazy to think that we have the capability, using statistics, to find out if X can be related to Y or Y related to X even if it is absurd. One of the things found later in the study was that Head Start, an early entry schooling program, does not affect the childs test scores in the future. It does not prepare them for a better education. Regression Analysis can be used for anything which is what surprises me the most.
Hint: Use message, echo and results in the chunk options. Refer to the RMarkdown Reference Guide.