Bivariate regression analysis and Multivariate regression analysis

[Mary_am]

Hey guys, I urgently need help with a Rstudio project that is getting impossible. I am not very familiar with the software as this is unrelated to my studies, but as it is an important one-time project I thought to get extra help from the community itself.
Using the European Value Study data-set (EVS_UK_full), I am measuring two variables:

• Dependent variable: attitudes towards immigration (are immigrants good or bad for the development of the country?) with answers going from 1 to 5, where 1 is very bad and 5 very good.
• Independent variable: income of respondents per year (household) with answers going from 1 to 10, with 1 representing respondents with less than 11,619£ a year and 10 with more than 64,070£ a year.
  I have done all the basic measurements to calculate the central tendency, the causal relationship. However, I don't seem to be familiar with testing the hypothesis that: a low level of income is linked with a negative attitude towards immigrants, while a high level of income generates a positive attitude towards immigrants. Apparently, to do so, I will need to go through the right statistical test, the bivariate hypothesis testing (tabular analysis - chi-square, difference of means - t-test, correlation coefficient, linear regression, or logistic regression, depending on the type of variable). After that, I will need to do the actual regression analysis and the multivariate regression analysis, which I really am not familiar with. I am just seeking for help in terms of the steps I have to take, theoretically speaking. For instance, given the variables stated above, what steps are best to measure the hypothesis given all the types of test to be done, what kind of variables are the two stated above? categorical, ordinal or interval?

Please let me know! Hope it is all clear and looking forward for this project to reach an end! As someone absolutely not familiar with coding, it really does get frustrating.

[Leon]

@Mary_am

As someone absolutely not familiar with coding

Hi @Mary_am,

R is extremely powerful, but you have to put in some time and energy in learning the basics and this does indeed include coding.

Given the project you describe, I suggest these two free ressources:

• R for Data Science
• Statistical Inference via Data Science - A ModernDive into R and the tidyverse

Hope it helps and good luck on your journey. You'll discover that the yield is well worth the energy input
Leon

[Mary_am]

Thank you Leon! I will dive into it right now.
It is absolutely challenging, however when it comes to assigning a category to variables or identify the right way of proceeding (like the bivariate regression analysis), it does get very frustrating, and often without a guide in how to proceed, it gets almost impossible. Especially given the fact that coding has so many rooms for interpretation. There isn't just one code to execute an action, there are so many!
All the best.

[Leon]

The reason, that you find coding so frustrating is that when you learn to program for the first time, you are in fact learning two things in parallel: i. what you actually have to type to do something (the syntax) and ii. what programming actually is, i.e. what does it mean to code and how do you translate problems into code (programmatic thinking).

Using the analogy of building a house, I see many being impatient and jumping straight to how-to-build-a-bathroom without realising, that you first need build a foundation, then the walls, etc.

Put in the initial energy of learning the subject from the bottom and you'll be plenty rewarded moving forward