Chapter 9
Chapter 10
Application
Chapter 12
100
A Chi-Square test is used for what kinds of variables?
Categorical
100
What is alpha level?
A threshold of probability below which we interpret our results as not being due to sampling variation
100
How would you describe the relationship in this scatterplot?
Weak, negative
200
What makes up the inner cells of a contingency table?
The frequency, or number of cases at that level of the independent and dependent variable
200
What is type 1 error?
Rejecting a null hypothesis when in reality the null hypothesis is true
200
Generate a question for a Pearson correlation coefficient test
Two ratio level variables
200
What does the slope or beta coefficient in the Least Squares regression line tell us?
How much Y increases with every one unit increase of X
300
The underlying mathematics of the Chi-Square test compares the observed frequencies in our table to what?
Expected Frequencies
300
How would we change our alpha level to limit type 2 errors?
Theoretically find the difference between our two sample means repeatedly and then create a probability distribution
300
Generate a question for a two sample t test.
Testing any two samples against each other to determine if they have the same population (if we don't have previous population data)
300
The numerator in the Pearson Correlation Coefficient Equation represents ---- and the denominator represents ----
The covariation between our variables
The variation within each of our variables
400
What is the difference between significance and effect size?
Significance refers to whether or not we think there is either a difference between our sample and population, two samples or if the IV has an effect on the DV. Effect size tells us how large that difference actually is. For instance, the fish in one river might be objectively larger than the fish in a different river, but the size of that difference may only be an inch or two.
400
If my sample means come from the same population, as I increase my sample size what would I expect to happen to the mean difference?
The means would become closer together (because they will better represent their shared sample population mean)
400
Generate a research question where I would use a Z score for my probability testing
A scenario where I am comparing my sample against a population that I previously had data for
400
What is a conditional mean of y and how is it used for creating our Least Squares regression line?
It is the mean of Y at any value of X. Our Least Squares regression line will come as close to each conditional mean of Y as possible, and this will lead to a OLS with the least amount of residuals or differences between our observed values and our OLS line
500
What will give us a clearer picture of difference between our variables, raw or percentage differences/frequencies? Why?
Percentage differences give us a better picture. They help to compare across our dependent variable within each level of our independent variable
500
if my sample size increase, but my mean difference stays the same, would I be more or less likely to say there is a statistically significant difference? How come?
More likely. Because we would expect the means to become closer together if they shared a population mean. Also, if they did share the same population mean it would be less likely to find the mean difference with a larger sample size.
500
Generate a research question for a question on confidence intervals and what data would I collect?
Mean,S and N
What is the average size of beavers in washington?
500
Why can't we use just the Pearson Correlation Coefficient, or just the Least Squares Regression Line? Why do we need both?
Pearson only tells us the strength of the relationship. It gives us know predictive ability's and is not in the units of the study.
The OLS can give us predictive power and uses appropriate units, but does not tell us about the strength of the relationship