Exam 2 Two-Way tables

Question 0

Bias

Answer 0

The design of a statistical study is biased if it is systematically favors certain outcomes.

Question 1

Association

Answer 1

Values of one variable tend to occur with certain values of another variable; detected when the conditional distributions differ from the marginal distributions and from each other.

Question 2

Bivariate data

Answer 2

Data collected on two variables for each individual in a study.

Question 3

The only arithmetic operation that makes sense for categorical data is

Answer 3

Counting

Question 4

Two-Way table is a

Answer 4

Table giving counts (or %) for two categorical variables

Question 5

In two-way tables the ….. Is is a row variable, and …… Is the column variable

Answer 5

Explanatory ….response

Question 6

Marginal distribution

Answer 6

Distribution of column variable separately ( or row variable separately ) expressed in counts or percent

Question 7

Both margins sum to the

Answer 7

Same overall total

Question 8

Conditional Distributions

Answer 8

Distribution of percents under a specific condition ( either in row or column)

Question 9

There is no association between categorical variables when all

Answer 9

Conditional distributions look like the marginal distribution and each other

Question 10

There is a potential association between categorical variables when

Answer 10

One or more of the conditional distributions look very different from the marginal distributions or each other

Question 11

Simpson’s paradox happens

Answer 11

When the third lurking variable is associated with the other two.

Question 12

What is Simpson’s paradox

Answer 12

A situation where the relationship between two variables is in one direction when the third lurking variable is included but in the opposite direction when the the third lurking variable is ignored.

Question 13

Measures affected by outliers

Answer 13

Mean
Standard deviation
Correlation
R-sq
Slope and y-intercept

Question 14

Rules of Data Analysis

Answer 14

1. Always plot your data
2. Always describe shape, center, spread of distributions
3. Don’t use the normal distribution to mod data that is not normal
4. Always draw and label the normal curve when finding % of x values
5. For Bivariate data, always examine a Scatterplot before computing r or modeling with a least squares regression equation