exam 2

Question 1

explain mean / pros+cons of it

Answer 1

mean = average
pro = quick idea about the data
con = sensitive to extreme values

Question 2

explain median / pro

Answer 2

median = middle value of distribution
pro = immune to extreme values

Question 3

explain mode / pro

Answer 3

mode = frequency of most recurring value
pro = gets an idea as to which answer is most popular in dataset

Question 4

standard deviation / what will get you a bigger standard deviation

Answer 4

looks at how much on average is each value in the data set differing from the mean

if people are far from the mean it will be a bigger standard deviation

Question 5

calculate range / what does a large range show about the data set

Answer 5

max value in data set - min value

a large range shows that people are giving low and high values

Question 6

what type of hypothesis testing should you use in this example: On a scale from 1-7 how much do you like the brand?

Answer 6

one sample t-test

Question 7

why would you use a one sample t-test?

Answer 7

when you are only dealing with one variable, it allows you to get some insight on just one

Question 8

how do you analyze results from a one sample t-test?

Answer 8

by comparing the midpoint of the scale (ex: 4 on a 1-7 scale) with the mean of peoples responses to see if people fall away from midpoint or not

Question 9

what types of hypothesis testing techniques do we use if we are dealing with 2 variables in cause and effect relationships?

Answer 9

multivariate techniques

Question 10

what type of hypothesis testing should you use in this example: Are students or faculty more likely to use the internet?

what are the dependent and independent variable(s)?

Answer 10

independent sample t-test

independent variable = students or faculty

dependent variable = how much they use the internet

Question 11

when would we use an independent sample t-test?

Answer 11

use when comparing two groups of responses to each other

Question 12

what type of hypothesis testing should you use in this example:

Hypothesis: in communities where air pollution is a problem (vs not a problem), consumers are more likely to purchase an EV.

what are the independent variable(s)? dependent variable(s)?

Answer 12

independent sample t-test

Independent = air pollution presence (categorical because just answer yes or no)

dependent = likelihood of purchasing an EV (continuous because it could be measured on a 1-7 scale)

Question 13

when running an independent sample t-test, how should i analyze the results? step by step

Answer 13

1. look at “two sided p value” section. is it less than 0.05? if so, we can conclude that there is a significant difference between the two groups
2. find the mean. which one is less? if the p value was significant, the lower value is more likely to do whatever it is

Question 14

what type of hypothesis testing should you use in this example:

how much do you like strawberry ice cream? how much do you like vanilla ice cream?

Answer 14

paired sample t-test

because we are comparing strawberry and vanilla on the bases of two continuous scales because this SHOULD be 1-7 scale

Question 15

when should you use an independent sample t-test ?

Answer 15

when we compare 2 groups of responses to each other

Question 16

when do we use a paired sample t-test?

Answer 16

When we need to compare means of two scores

when we need to compare two variables in the data which should be measured on a continuous scale

Question 17

what type of hypothesis testing should you use in this example:

lets predict people purchase an EV to help the environment more than to save money on gas.

what questions should we ask?

Answer 17

paired sample t-test

questions: 1. measuring whether people purchase an EV to save on gas 2. measures whether people purchase an EV to help the environment

Question 18

how do we conduct a paired sample t-test analysis? step by step

Answer 18

1. look at if one mean of one variable is higher than the other
2. confirm the difference by looking at p value. is this less than 0.05?
3. if so, we can confirm that the lower value of means = more likely to do that thing

Question 19

when do you use an ANOVA? what types of variables to the X and Y need to be?

Answer 19

use when you compare two or more groups. your independent variable has to be categorical and dependent variable needs to be continuous

Question 20

what type of what type of hypothesis testing should you use in this example:

different age groups and their frequency is using disney plus subscriptions

why?

Answer 20

ANOVA because you are measuring a categorical independent (age) with a continuous dependent (frequency of use)

Question 21

how do you analyze an ANOVA table to get results? step by step

Answer 21

1. look at the means. are there any categories that have higher means than others?
2. look at significance. is it below 0.05? if so, we see that there is a difference between the groups.
3. if there is a difference between groups means, you need to prove it.
4. look at follow up Scheffe and Ad Hoc tests
5. find the “significance” section under the multiple comparisons table. are there any significance values below 0.05? if so, we can conclude that there is a significant difference between those groups. if the sig value is over 0.05, then there is no sig difference between those two groups

Question 22

explain linear relationships

Answer 22

association between two variables wherein the strength and nature of the relationship remains the same over the range of both variables

Question 23

explain curvilinear relationships

Answer 23

a relationship between two variables wherein the strength/direction of the relationship changes over the range of both variables

Question 24

when would we use a correlation?

Answer 24

when we are examining the relationship between two non categorical variables (through interval or ratio variables)

Question 25

what are three things to look for when assessing correlation/regression?

Answer 25

1. does the relationship exist 2. what direction is the relationship in? (Negative or positive) 3. strength of the relationship

Question 26

when using correlation/regression, how can we tell if a relationship exists ?

Answer 26

if the sig value is less than 0.05, there is some relationship here

Question 27

how can we find out the direction of the relationship in a correlation/regression?

Answer 27

look at the sign of the coefficient if it is positive, then x and y are both increasing if it is negative, either y is increasing while x is decreasing or vice versa

Question 28

how do you discover the strength of a correlation/regression relationship?

Answer 28

through the value of the coefficient if the pearson correlation coefficient is closer to 1, the relationship is stronger. if it is closer to 0, the relationship is weak

Question 29

what is the pearson correlation coefficient?

Answer 29

measuring the relationship between two continuous variables a statistical measure of the strength of a linear relationship between two continuous (interval or ratio) variables

Question 30

what values does the pearson correlation coeff. go between?

Answer 30

Value is going to vary between -1 and +1

Question 31

what does a pearson correlation coeff. of +0.8 mean?

Answer 31

the relationship between two continuous variables is strong

Question 32

what does a pearson correlation coeff. of +0.2 mean?

Answer 32

there is a weak positive correlation between two continuous variables

Question 33

what does a pearson correlation coeff. of -0.5 mean?

Answer 33

-0.5 indicates a moderate negative linear relationship between two continuous variables.

Question 34

given this example, what test would we run and what steps would we need to take in order to analyze if the satisfaction with the service is related to a cosumer's likelihood of recommending it? : suppose we are looking at the disney plus video streaming service and trying to assess whether the satisfaction with the service is related to a consumers likelihood of recommending it.

Answer 34

we would run a pearson correlation coefficient because these are both continuous variables steps: 1. look @ sig value. is it less than 0.05? if so, there is a relationship between the variables 2. find the pearson correlation coeff. is it positive? if so, we can say that there is a positive relationship between the variables 3. is the relationship strong? we will assess this if the coeff. is closer to 1. if it is closest to -1 then it is not strong.

Question 35

when do you use substantive sig? what type of coefficient do we look at to check for substantive sig?

Answer 35

if there was a weak and significant relationship in a correlation/regression. Look at r^2 (or the coefficient of determination) to check for substantive significance

Question 36

what is r^2 called?

Answer 36

coefficient of determination

Question 37

if the coeff. of determination is large, what does this say about the linear relationship?

Answer 37

The larger the size of the coefficient of determination, the more meaningful the linear relationship between the two variables being examined is

Question 38

what would you say about the meaningfulness of the relationship between variables if r^2 = 0.60?

Answer 38

if r ^2 = 0.6, we would say that 60% of the variation in one variable is associated with the other variable of interest. So the relationship would be meaningful

Question 39

what is r^2 the square of?

Answer 39

the pearson correlation coeff.

Question 40

what range does r^2, or coeff. of determination, vary from?

Answer 40

0.0 - 1.00

Question 41

r^2 def

Answer 41

a number measuring the proportion of variation in one variable accounted for by another

Question 42

if r is 0.8, what is r^2?

Answer 42

0.64

Question 43

if you have an r^2 of 0.64, is this relationship of substantive significance practically useful? explain

Answer 43

Since we got 0.64 = 64%, this means that 64% of the variation in one variable is being explained by the other variable. This is pretty strong which means that one variable (independent) is having a pretty large influence on the other (dependent)

Question 44

if i had a value of 0.2 for my correlation coefficient, what would be the r^2 value? what does that value mean? is this relationship of substantive significance practically useful?

Answer 44

0.2 x 0.2 = 0.04. This means that only 4% of the variation in one variable is being explained by the other. Therefore, the weaker the correlation, the smaller the amount of variance in one variable can be explained by the other variable. example: Even if i make people super satisfied with my product, its only going to have 4% of the impact on their likelihood to recommend.

Question 45

Correlation ____ completely tell us whether there is a cause and effect

Answer 45

cannot

Question 46

explain x and y relationship with a correlation

Answer 46

in a correlation, x is associated with y

Question 47

explain x and y relationship with temporal priority

Answer 47

cause must come before the effect, x must come before the y ex: studying whether using disney + more often (x) causes high satisfaction (y). to prove this, you must show how people started using disney + more before they reported higher satisfaction

Question 48

explain x and y relationship with a non-spuriousness

Answer 48

the correlation between x and y is not a result of an outside confounding variable

Question 49

why use regression? what does it tell me?

Answer 49

it predicts values. if i change one variable, it tells me how much of a change i can expect in the other

Question 50

make a regression out of this: Example: say we were looking at something like the effect of price on satisfaction

Answer 50

If i change price by $1, how much can i expect satisfaction to change?

Question 51

with regressions you need to focus on ___

Answer 51

slope

Question 52

what does slope tell you?

Answer 52

the relationship between x and y

Question 53

what does a flat slope mean in a linear regression?

Answer 53

means that if I were to change x by a lot, y is not going to change that much. The flatter the line, if i keep going towards the right, and keep increasing values of x, y will barely change.

Question 54

what does a steep slope mean in a linear regression?

Answer 54

As the line becomes steeper, the influence of x on y starts to become large So whenever you see higher values for the slope, it means that x is having a very large influence on y

Question 55

the steeper the slope, the larger the effect _ has on _

Answer 55

x has on y

Question 56

why is line of best fit good? what does it tell you

Answer 56

For all of your data points, you want to minimize the distance between the data point and the line that youre drawing between them

Question 57

what kind of test should i run on this example? How the square feet of a house affects the price of a house what are the independent / dependent variables?

Answer 57

X independent variable = square feet Y dependent variable = price linear regressions

Question 58

steps for analyzing linear regression table:

Answer 58

go to coefficients table 1. find sig value. is it less than 0.05? if so, we can conclude that x (independent) has a significant effect on y (dependent) 2. find b coeff. is it positive? if so, we can conclude that if we were to increase the independent variable by 1 unit, it should result in an increase in the dependent variable 3. look @ b value again. if you were to increase ind. by 1 unit, change in dep. variable will be the b value

Question 59

what type of testing do we do when we have more than one ind. variable in your regression?

Answer 59

multiple regression

Question 60

steps for analyzing multiple regression tables

Answer 60

1. look @ ANOVA table. is the sig value less than 0.05? if so, the regression is significant. 2. go to coefficients table. find the sig values for both independent variables. are they less than 0.05? if so, then they are significant. if they are not below 0.05, then the ind. variable does not have an effect on the dependent variable.

Question 61

Unstandardized Coefficients

Answer 61

they tell you how much the dependent variable changes when one independent variable increases by 1 unit, while holding other variables constant.

Question 62

when do we standardize our scales? give an example

Answer 62

When we have two variables that are on different scales and cannot necessarily be compared against each other (ex: we may have number of rooms AND square feet of the house which are not comparable)

Question 63

why do we standardize coefficients ?

Answer 63

allows us to see which one has a larger influence on the Y

Question 64

The larger the standardized coefficient, the larger the ___.

Answer 64

impact

Question 65

to reach more specific conclusions, do you want to look at standardized or unstandardized coefficients?

Answer 65

unstandardized

Question 66

when do we use Binary logistic regression

Answer 66

May use when independent variable is continuous but your dependent variable is categorical

Question 67

can you give an example of when we would use binary logistic regression?

Answer 67

Say we were looking at how price affects peoples tendency to purchase and purchase is measured as a yes or no. Continuous variable of price ranging between $100-$200 Categorical variable of tendency to purchase as only yes or no