MIS major

Correct

Was wrong

Question 1

According to the instruction in class, which of the following is true:

· The correlation coefficient between two variables DOES vary greatly depending on their underlying UNITS of measurement (ex: using grams vs kilograms scales changes the correlation between weight and height, etc)

· When determining the correlation coefficient r as part of a regression analysis it DOES matters which variable is considered to be the dependent variable and which the independent

· both of the above are true

· None of the above are true

Question 2

The r (correlation) values range from-----.

· -1 to 1

· 0 to 1

· 0 to 100

· 0 to infinity

Question 3

The r-square values range------.

· -1 to 1

· 0 to 1

· 0 to 100

· 0 to infinity

Question 4

Assume you ran a bivariate regression model (remember bivariate means there is only one independent variable) in which # of cans of tuna fish consumed per week was the independent variable (IV) and level of hair silkiness was the dependent variable (DV); the hypothesis being eating more tuna is associated with more silky hair (because, logically, sea lions eat more fish and they're more silky feeling so why not humans, too).

The model resulted in a correlation r =.6, beta coefficient = 2.14, and a P-value =.004.

If you switched the independent and dependent variables and ran a new bivariate regression analysis (so cans of tuna = DV and hair silkiness = IV), which of the following would substantially change?

· The correlation “r” value

· The regression coefficient

· The P-value

· All of the above would substantially change

· None of the above would substantially change

Question 5

A correlation of r = 0 indicates that

· X and Y don’t have a linear relationship

· X and Y are unrelated

· X and Y a linear relationship

· None of the above

Question 6

According to the professor, if we find the P-value associated with an unstandardized regression coefficient = .05, then the associated 95% confidence interval:

· is the range of the effect across 95% of the population

· indicates we are 95% confident that the coefficient is accurate

· both of the above are true

· None of the above are true

Question 7

According to the instruction on statistical significance in class, the confidence level is the probability that a confidence interval will include the population parameter.

· True

· False

Question 8

Which of the following is true?

· If a difference in measurement is statistically significant, then it is also practically significant

· If a difference in measurement is practically significant, then it is also statistically

· Significant

· both of the above are true

· None of the above are true

Question 9

According to the discussion on sample size and statistical significance (hint: think of the excel sample size calculator), when calculating sample size

· You need to provide the estimated margin of error

· You need to provide the Z score (or the corresponding P- Value)

· both of the above are true

· None of the above are true

Question 10

According to the discussion on sample size and statistical significance (hint: think of the excel sample size calculator), when the level of variance increases, the sample size needed to maintain a particular level of statistical significance

· Increase

· Decrease

· Is not affected (can main the same)

· Can’t be determined

Question 11

As described in class, when it comes to parallel computation of regression analysis on very large datasets (in which the data is spread over many machines, analysis is run on each machine, and then averages are taken across all of the analyses), the resulting averages are good approximations for which of the following computations (assuming no extra data manipulation):

· regression coefficients

· 95% confidence intervals

· both of the above

· none of the above

Question 12

As described in class, when it comes to parallel computation of regression analysis on

very large datasets (in which the data is spread over many machines, analysis is run on

each machine, and then averages are taken across all of the analyses), the resulting P-value scores (assuming no extra data manipulation) are --------- what they would be if the analysis of the dataset was performed on a single supercomputer

· good approximations of

· larger than

· smaller than

Question 13

One of the advantages of very large datasets (big data) is that we no longer have to worry about multicollinearity among independent variables in regression analysis

· True

· Fales

Question 14

Multicollinearity is problematic in a typical sized dataset (ex: couple hundred observations and several variables) if high correlation (above .85 or .90) exists between:

· Two or more independent variables used in the multivariate regression model

· Any of the independent variables and the dependent variable used in the multivariate regression model

· both of the above

· none of the above

Question 15

Factor analysis can be used in which of the following?

· To identify underlying dimensions, or factors, that explain the correlations among a set of variables

· To identify a new smaller set of uncorrelated variables to replace the original set of correlated variables in subsequent multivariate analysis.

· To identify a smaller set of salient variables from a larger set for use in subsequent analysis.

· All of the above are correct circumstances.

Question 16

According to the professor, including all of the criterion variables related to the managerial decision under question (aka including every variable we are interested in that is in the database) in cluster formulation will IMPROVE the insight of cluster analysis.

· True

· False

Question 17

Which of the following is NOT one of the 3 requirements of market segmentation?

· Identifiable

· Reachable

· Sizeable

· Loyal

Question 18

The three benefits of market segmentation include

· Identifies opportunities for new product development

· Improves the strategic allocation of marketing resources

· both of the above

· none of the above

Question 19

Consumer Market Major Segment Bases include all of the following EXCEPT:

· Demographic

· Psychographic

· Teleographic

· Behavioral

Question 20

In the usage based approach to market segmentation, every company at least _______segments:

· One

· Two

· Three

· Four

· Five

· Eight

Question 21

Cluster analysis does not classify variables as dependent or independent

· True

· False

Question 22

When it comes to linking methods in cluster analysis, which of these statements are true?

· each of the linkage methods can yield different results when used on the same dataset

· each linking method has its specific properties

· both of the above are true

· none of the above are true

Question 23

Which statement is NOT true about cluster analysis?

· Cluster analysis is a technique for analyzing data when the criterion or dependent variable is categorical and the independent variables are interval in nature.

· Cluster analysis is also called classification analysis or numerical taxonomy.

· Groups or clusters are suggested by the data, not defined a priori.

· Objects in each cluster tend to be similar to each other and dissimilar to objects in the other clusters.

Question 24

When it comes to K-Means clustering, which of the following is true?

· we do not need to tell it how many clusters there are to begin with

· we can assess the best configuration by examining the dissimiliarly coefficient in the K-means generated agglomeration schedule

· both of the above are true

· none of the above are true

Question 25

When it comes to K-Means clustering, which of the following is true?

· we can save the cluster membership

· using different combinations of variables can result in different cluster assignments for each observation

· both of the above are true

· none of the above are true

Question 26

When it comes to very large datasets (i.e., a form of big data) and cluster analysis, it was recommended in class to:

· just use hierarchical cluster analysis

· just use k-means cluster analysis

· use both hierarchical and k-means cluster analysis together

· never use cluster analysis

Question 27

Clustering can be performed using:

· Observable (directly measured) variables such as dollars spent on different products, etc

· Unobservable (inferred) variables measured on surveys such as attitudes and moods etc. (ex: happiness, sadness, etc. liked scales)

· both of the above

· none of the above

Question 28

If you switch the dependent and independent variables in a regression analysis, which of the following will usually change:

· Regression Coefficient Value

· P-value

· both of the above change

· none of the above change

Question 29

If you switch the dependent and independent variables in a regression analysis, which of the following will usually change:

· Regression Coefficient standard error

· Correlation (r) score

· both of the above change

· none of the above change

Question 30

The 3 types of linkage in cluster analysis include:

· single linkage

· complete linkage

· both of the above true

· none of the above true

Question 31

If we have several hundred independent variables that might be related, which technique do we use to reduce the number of variables?

· Factor Analysis

· Cluster Analysis

· Compare Means Analysis

· Numerical Taxonomy

· Classification Analysis

Question 32

When running Factor Analysis, which of the following is recommended by the professor:

· We do not rotate the data

· We select a Promax rotation

· I We select a Varimax rotation

· We select a Quadmax rotation

· We select a Expert rotation so SPSS will decide which rotation is best

Question 33

When running Factor Analysis, we select to suppress absolute value smaller than

· .01

· .05

· .1

· .4

· .95

For questions 34 to 36, assume that you ran a regression to predict which

students are most likely to agree to dress up as a circus clown for $5 at a charity baseball game at the local elementary school. The results are:

Dependent Variable: Likelihood of Dressing Up as a Clown (Very Unlikely 1, 2, 3, 4, 5, 6, 7 Very Likely)

Question 34

Is someone with red hair more or less likely to dress up as a clown than someone with blue hair?

· More likely

· Less likely

· Equal likelihood

· It is not possible to determine from the provided data

Question 35

Referring back to the regression coefficients from question 34....

Is the coefficient for “Age” statistically significant at a 95% confidence interval?

· Yes

· No

· It can’t be determined from the provided information

Question 36

Referring back to the regression coefficients from question 34....

Is the coefficient for 'likes watching Dr. Who' statistically significant at a 95%

confidence interval?

· Yes

· No

· It can’t be determined from the provided information

Question 37

Referring back to the regression coefficients from question 34....

Who of the following is most likely to dress up as a clown at the charity?

· ride a moped, has red hair, 20 years old

· rides a bicycle, has blue hair, 20 years old

· like watching dr who, has blue hair, 20 years old

Question 38

When it comes to the exponential smoothing model in forecasting, which following are true:

· The model has received widespread acceptance among American business firms that employ sales forecasts for managerial planning and control

· It uses special weighted moving averages and seasonal factor that is multiplied by the weighted moving average to calculate the forecast.

· both of the above are true

· none of the above are true

Question 39

Generally, the exponential smoothing model uses _________ smoothed statistics that are weighted.

· Zero

· One

· Two

· Three

· Four

· Five

Question 40

One of the most widely used techniques for short-term forecasting that is autoregressive integrated moving average (ARIMA) model associated with G. E. P. Box and G. M. Jenkins

· True

· False

Question 41

Which of the following is true regarding the Arima forecasting technique?

· It is somewhat mathematically tedious and complex

· It relies of using past sales data exclusively

· Both of the above are true

· None of the above are true

Question 42

In excel, the forecast sheet button produced a graph and a set of forecast estimates. According to the professor, this techniques uses which forecasting approach?

· simple moving average

· exponential smoothing

· ARIMA modeling

· ARMA modeling

Question 43

In excel, if you have five columns each with a different year of data (think 2015, 2016, 2017, 2018, 2019) the spreadsheet has to be sorted so the oldest/smallest columns (ex:2015) always are to the left and newest/largest values always to the right (ex: 2019) or the forecast can’t ever calculate the next predicted value correctly.

· True

· False