INST314 review 3

To meet the assumption of multivariate normality, the residuals of the model must be normally distributed

Projects 2 and 3 are group projects

The specific type of generalized linear model used for Poisson data is

Which of the following is true of resampling methods

If two variables have a strong negative correlation, Pearson's r will be

Multi-select: Statistics and machine learning both use statistical probability models, but in statistics scientists

Regression models are valid for data outside the range they were developed for

The assumption of homoskedasticity for a regression model implies that the model performs equally well throughout

The non-parametric alternative to the ANOVA for numeric or ordinal data that tests median instead of mean is called

Correlation analysis is not strongly impacted by outliers

There is an in class final for this class

If a p-value is larger than alpha we reject the null.

Bootstrapping will allow a researcher to extrapolate results to sub-populations that were not in the sample

An ANOVA will tell you if one or more means has a statistically significant difference, but not which is different.

The null hypothesis for a binomial exact test is

The chi-square test of independence

ChatGPT is an example of a machine learning model

To compare the means of 2 parametric samples that are not independent, you would use

Project 3 is due on May 19th