So imagine my chagrin/relief when I was finally forced/able to take a research methodology course that included discussions of quantitative methodology and statistical analysis. It's like a general entry into the world of statistics, but not as scary as a full on stats course. Over the next few posts, I will write about some of the things that I learned about quantitative research methods, which is fortuitous timing because I have to study for the final exam.
First things first, I basically had to re-learn basic science concepts. It turns out that my grade 3 teacher lead me astray on the basic science concept, variables, when we were growing peas in science class and she said that the independent variables were the things that you manipulated to see how the plant would grow: like sunlight, water, soil, etc. Then she said that the dependent variable is the thing that you don't change - like the pot that the plant is in. WRONG. The dependent variable is the plant. It is the thing that changes as the independent variable is manipulated.
I basically had to unlearn everything I know about science and then re-learn it thanks to her.
In addition to the independent variable and the dependent variable, there is also an intervening variable. The intervening variable is changed by the independent variable and then the intervening variable turns around and causes change in the dependent variable.
For example, we've all heard that people with higher levels of education have higher incomes. However, this can't quite be true because I have 7 years of university education under my belt and currently my salary is $0.00 because I don't have a job. The intervening variable here is occupation and the idea is that when you're more educated, you can get a better job, which results in a higher income.
For example, we've all heard that people with higher levels of education have higher incomes. However, this can't quite be true because I have 7 years of university education under my belt and currently my salary is $0.00 because I don't have a job. The intervening variable here is occupation and the idea is that when you're more educated, you can get a better job, which results in a higher income.
Lastly, there are external, extraneous, or confounding variables which the scientist can't really control for but may still have an impact on the experiment. These are also called uncontrolled variables.
Another concept that I missed/forgot about entirely is the hypothesis. I remember it being a statement of what the scientist predicts is going to happen, which is somewhat true, but what no one ever mentioned to me is that hypotheses are always in pairs: the null hypothesis and the alternate hypothesis.
The null hypothesis is always stated in negative terms. "The independent variable will not affect any changes in the dependent variable", "the heat setting on my curling iron will have no effect on how long my curls last", "Miley Cyrus' cd sales are not related to Iron Maiden's cd sales".
The null hypothesis is paired with an alternate hypothesis, which is basically just the null hypothesis stated in exactly the opposite terms: like reverse psychology, or what I would say in response to everything my brother ever said ever.
"The independent variable will affect changes on the dependent variable", "the heat setting on my curling iron will have an effect on how long my curls last", "Miley's cd sales are related to Iron Maiden's cd sales".
Why is it important to know about hypotheses and variables for statistics? Because when doing statistical tests with quantitative data, you have to have variables, otherwise you would have nothing to test. You also have to have a hypothesis otherwise you won't really know what your test results are saying.
Why is it important to know about hypotheses and variables for statistics? Because when doing statistical tests with quantitative data, you have to have variables, otherwise you would have nothing to test. You also have to have a hypothesis otherwise you won't really know what your test results are saying.
Next up...what exactly is a beta coefficient and other things to know when reading published statistical research.
No comments:
Post a Comment