2:Descriptive Stats Reslationships Between Varibles Part 2

Question 1

Pearson r

Answer 1

Two quantitative variables
Correlation coefficient
-linear relationship between 2 QUAN variables (interval or ratio) normal
Sign of r tells u direction of relations between variables
Magnitude of r absolute value tells u the strength of linear relationship between variables

Question 2

Cohens d

Answer 2

One dichotomous categorical and one quantitative to compare means of two distributions

Question 3

Scatterplot

Answer 3

X= IV
Y= DV
Display paired data from a set of individual cases
-may include best fit

Question 4

Positive correlation

Answer 4

Increase in one is increase in other. Positive slope

Question 5

Negative correlation

Answer 5

Increase in one variable accompanies a decrease in other. Negative slope.

Question 6

Magnitude of r

Answer 6

The larger the magnitude of r, the closer the data points are to the best-fit straight line.

Coefficient determination: r^2 percentage of variation in B explained by A

Question 7

Side by side box plots

Answer 7

Comparing means. Data from multiple

Groups or conditions compared side by side.

Question 8

Bar graphs

Answer 8

Comparing means.
Display means as bars extending from categorical axis
-heigh/length of each represents mean value of quantity variable for cases at that value of categorical variable
-error bars illustrate variability as whiskers extending from top of bar
EMPHASIZE DIFFERENCES BETWEEN MEANS

Question 9

Factorial designs

Answer 9

One categorical variable along the categorical axis (X)

-other categorical variable in legend

Question 10

Line graphs

Answer 10

Comparing means
Means as points connected by line segments
-x is ordinal scale or higher
-if x variable is quantitative but discrete, the x axis is scaled with units (distance is meaningful)
-EMPHASIZE CHANGES IN MEAN VALUE OF Y AS FUNCTION OF X

Question 11

Functional relationships

Answer 11

Characterized based on how the value of Y changes as the X increases
Monotonic: continuously increasing or decreasing
Nonmonotonic: contains at least one reversal in direction
Asymptotic: if curve levels off

Question 12

Cohens d effect size

Answer 12

Difference in means relative to pooled standard deviation of 2 distributions
-less overlap of distributions = larger cohens d = larger effect size
(Stronger relationship or association)

M1-M2/ SD

Question 13

If both variables categorical

Answer 13

Must display counts or proportions for number of cases at each combo

• frequency table
• multicolor or patterned histogram with two or more series
• side by side pie graphs