Determine which statistical test to use.

This is a multi-step question with several components, going through the process of selecting a statistical test. I am genuinely trying to understand this process and it would be very helpful if details were provided on what I have labeled incorrectly.

The variables are: highest level of education and quality of life

Data is being pulled from GSS Data Explorer.

1) Do I have the independent and dependent variables identified correctly? If not, what is wrong?

Independent Variable (presumed cause): highest level of education (high school - graduate).

Dependent Variable (presumed effect): quality of life (rated as poor - excellent)

State the question for hypothesis:

Does having a higher level of education {effect, impact, or improve} quality of life?

2) Should the question be simplified, why?

Does level of education {effect, impact, or improve} quality of life?

3) Does it make a difference which word {effect, impact, or improve} I use? Will it change the meaning or direction of the question and hypothesis?

4) Which null hypothesis and alternative hypothesis is written correctly and why?

H0: highest level of education is equal to quality of life

H1: highest level of education is not equal to quality of life

H0: level of education significantly {effects, impacts, or improves} quality of life

H1: level of education does not significantly {effect, impact or improve} quality of life

5) Do I have data labeled correctly?

Data: level of education is nominal, categorical, qualitative

*This is where I am unclear.

is quality of life ordinal, qualitative.

or is quality of life quantitative because it is rated on a scale?

6) Do I have the sample labeled correctly?

Sample: 2 separate or independent groups. Highest level of education and quality of life.

*This is where I am unclear. I have one independent variable and one dependent variable.

Does that make this 2 independent samples?

An independent and a dependent sample?

2 dependent samples?

7) Do I have the purpose labeled correctly? Purpose: to find the relationship between level of education and quality of life

(does level of education {effect, impact, or improve} quality of life

*This is where I am unclear.

I do not believe I am comparing variables or am I because I used "equal" in the first hypothesis set?

Am I looking for a correlation between two variables, instead of a relationship? or are they the same?

8) Which statistical test is correct?

*This is where I am unclear.

I believe I need a non-parametric test because the data is both nominal and ordinal. I don't think the data will be normally distributed, I think there will be outliers, and most likely unequal variances.

However, the sample sizes are much larger (in the thousands) than the recommended 30 for a nonparametric test.

If this is 2 independent samples, then I would need a Mann-Whitney U test. But I am not trying to determine if the dispersion of ranks in two independent groups are equal, I believe I am looking for the relationship.

If this is a correlation between two variables, then I would need a Spearman-correlation. But Spearman correlation requires the data to be ordinal, only one of my variables is ordinal.

Chi-square is used for categorical data, but requires an expected value to compare the observed value. I do not have expected values with the data presented, nor do I know how to find expected data for these samples.

One flow chart suggested a linear regression, but I have 2 categorical variables. The linear regression requires interval data and more variables.

I appreciate the help and feedback for this question. Because I am trying to understand the process of each step, I may have follow up questions after the initial response.