Excel's __________ can be used to construct a crosstabulation

Excel User’s Manual with Exercises

for Marketing Research, 9th Edition

Please note: Wiley is no longer distributing SPSS with our textbooks. IBM, owners of SPSS, is no longer partnering with any publisher to distribute SPSS with any textbook. This is a decision made by IBM and is outside of Wiley’s control. Access to SPSS can be purchased by going to this website http://www.ibm.com/software/analytics/spss/products/statistics/gradpack/ and by clicking on “Buy Student Versions” in the bottom right corner. If you have questions you can contact:

V. Monica Young

Sales Specialist, Publications

Business Analytics

Tel: (312) 651-3157

E-mail: monicay@us.ibm.com

PREFACE

This Excel User Guide is provided for students who will be working with the Excel version of the SPSS Exercises to accompany McDaniel and Gates, Marketing Research, 9e. It consists of detailed step-by-step instructions for each exercise, accompanied by relative visual aids, such as “screen shots,” to further serve as “road map” indicators ensuring that you are on the right track. The guide also includes periodic “troubleshooting” tips such as, “If you see “#####” in cell A3, then you need to…”

NOTE: All Exercises must be completed in chronological order, omitting none, for successful completion of this data analysis project.

NOTE: You must have the "Data Analysis Toolpack" installed for your Microsoft Excel program. To check to see if you have it: open Excel and click "Tools" on the toolbar at the top of the screen. If you see "Data Analysis" in the tools menu, then you are ready to go! If you do not see "Data Analysis" you will have to select "Add-Ins..." from the menu, click the "Data Analysis ToolPack" box, and hit "Ok".

If something happens whereby the computer can't find the file needed to install the pack, then ask your professor or lab assistant for help!

TABLE OF CONTENTS

Preface ii

Chapter 14: Sample Size Determination 1

Exercise #1: Sample Size Determination Using the Sample Means Method 1

Exercise #2: Determining the Reliability/Confidence of Sample Results 5

Chapter 15: Data Processing and Fundamental Data Analysis 7

Exercise #1: Machine Cleaning Data 7

Exercise #2: Analysis of Data with Frequency Distributions 11

Exercise #3: Analysis of Data with Descriptive Statistics 13

Exercise #4: Analysis of Demographic Characteristics Using Charts 15

Chapter 16: Statistical Testing of Differences and Relationships 18

Exercise #1: Analyzing Data Using Crosstabulation Analysis 18

Exercise #2: T/Z Test for Independent Samples 28

Exercise #3: ANOVA Test for Independent Samples 38

ii

Chapter 14 – Sample Size Determination

Exercise #1: Sample Size Determination Using the Sample Means Method

1

(1) Go to the Wiley website at www.wiley.com/college/mcdaniel and download the “Segmenting the College Student Market for Movie Attendance” database for Excel. Also download a copy of the “Segmenting the College Student Market for Movie Attendance” questionnaire so that you can understand the database contents. The most important items in the survey are in question #5. It contains 9 movie items in which respondents rate their relative importance. When you open the database, take note of the variable descriptions and the computer coding for each of the variables, which are contained in the “Variable Labels” and the “Value Labels” worksheets of the Excel Workbook.

(2) As you’ve learned in class and from reading the course text, the Sample Means method of sample size determination consists of:

i. required confidence level (z)

ii. level of tolerable error (e)

iii. estimated population variance (2)

iv. estimated sample size (n)

v. Formula: n = (z2 * 2)/e2

(3) Of the various methods of deriving sample size, estimated population standard deviation can be estimated based on prior studies, expert judgment, or by conducting a pilot sample. For this problem, we are going to estimate population standard deviation using a pilot sample . To do this you will use only the first 200 cases in the Segmenting the College Student Market for Movie Attendance database. We are assuming that these are responses to a pilot sample, and we will use them to estimate the needed sample size.

Do This : In the database, you’ll notice that questionnaire number 200 falls into line 201 because the field titles are in row number 1. So, highlight row “202” in the database, click “insert” on toolbar at top of screen, and select “row.” You do this so that you can have a blank working space in which to calculate the standard deviation of the pilot sample. (if you mess up and insert a row in the wrong place, just select the row, click “edit” and select “delete”; then insert the row in the right place).

(4) Now, Do This: select cell F202 and enter this formula: =STDEV(F2:F201)

This formula computes the standard deviation for all 200 responses for variable Q5a.

If you have #NAME instead of some decimal number in cell F202 of your database worksheet, then you forgot to put the colon (:) between the F2 and F201 in your formula. When you have something to the nature of 0.5 in your cell F202, then you’re good to go. Now select that cell and position the cursor until you have the fill tool indicator that looks like a cross. Now drag the tool cursor all the way to variable Q5i and drop it.

Congratulations, you just calculated the standard deviation for all of the components of question five of the survey (the most important questions that we are interested in, as stated above). We are assuming that each of the 9 variables is equally important with respect to the research objective.

(5) As you have learned from class lectures and from reading your course text, sample size determination can be reached by selecting the variable with the largest computed standard deviation.

Answer the following questions:

1. Which of the 9 movie theatre items had the largest standard deviation? ________________

2. Now, using the formula for the sample means method of sample size determination that you learned in class, make the necessary computations for each of the following:

a. Compute sample size given the following:

i. required confidence level (Z) is 95.44%.

ii. tolerable error (e) is .1 or 1/10 of a response point.

iii. standard deviation () = _____________________

iv. sample size (n) = ________________________

· Compute sample size given the following:

i. required confidence level (Z) is 99.72%.

ii. tolerable error (e) is .1 or 1/10 of a response point.

iii. standard deviation () = ___________________

iv. sample size (n) = ________________________

3. How do your computed sample sizes in the problems above compare to the total number of cases in the Segmenting the College Student Market for Movie Attendance database?

____________________________________

4. We are going to assume that the objective of our research concerning students attendance at movies can be expressed as a dichotomy (greater or lesser, etc.), for example, it doesn’t matter how much one group attends movies over another group, but just who attends the most. To accomplish this we can use the much less complicated sample proportions formula. We are going to assume that we have no prior studies, hence, in the sample proportions formula P = .5 and (1 – P) = .5. You will not need Excel to assist you with this computation .

a. Compute sample size given the following:

i. required confidence level (Z) is 95.44%.

ii. tolerable error (e) is .05 or accuracy within 5% of the true

population mean.

iii. standard deviation P = .5 and (1 – P) = .5

iv. sample size (n) = ________________________

b. Compute sample size given the following:

v. required confidence level (Z) is 99.72%.

vi. tolerable error (e) is .03 or accuracy within 3% of the true

population mean.

vii. standard deviation P = .5 and (1 – P) = .5

viii. sample size (n) = ________________________

NOTE: Once you have completed this exercise, DELETE the new row that you have created (row 202) before proceeding to the next exercise to avoid treating the standard errors as a row of data.

Exercise #2: Determining the Reliability/Confidence of Sample Results

(1) Instead of determining the needed sample size, we will now evaluate the confidence level of results derived from the entire Segmenting the College Student Market for Movie Attendance database. To evaluate this type of confidence, using the sample means formula, solve for Z instead of n. Hence, use the formula Z 2 = n * e2/o2. Then take the square root of Z2. You can go to the normal distribution table in the appendix of your text to determine the confidence level associated with the database. For the sample proportions formula, solve for Z using the formula Z2 = (n * e2)/[P(1 – P)] , then take the square root of Z2.

You can also use Excel to calculate the confidence level for the entire database. As you know from class, a confidence level for a normal distribution requires two tails. So you multiply 1-α by two. But don’t worry about that by itself… just enter the following formula into a blank cell, such as A503: =2*NORMSDIST(your Z value)-1.

A good Z-value is 1.75, since it corresponds to a confidence level of a little over 90% (the general cut-off for significance).

(2)

(3) Now, remember that we assume that question #5 has the most important questions in the questionnaire, with respect to the research objectives. Let’s calculate the standard deviation for these questions, but this time for the entire database instead of just the first 200 responses.

(4) Do This: select cell F504 and enter this formula: =STDEV(F2:F501) in order to have Excel compute the standard deviation for all responses for question 5a. It should be something close to 0.59.

REMINDER: You will need to DELETE row 202 from the prior exercise if you haven’t already. Otherwise, the calculation here will be off. The final line of data should be row 501 as shown above.

Now select that cell and position the cursor until you have the fill tool indicator that looks like a cross. Now drag the tool cursor all the way to variable Q5i and drop it. Now you have computed the standard deviation for all responses to Q5a through Q5i. Choose the question with the largest standard deviation and use it in the formula for computing the confidence level in the following problem.

(3) Given the preceding, compute the confidence level associated with the Segmenting the College Student Market for Movie Attendance database, given the following:

1. a. tolerable error is .1 or 1/10 of a response point.

b. sample size = 500.

c. standard deviation ______________________

2. Confidence Level = _______________________%

3. How do the results in 2. above compare to the results in 2. of the sample size determination problem?

___________________________________________________________

(4) Sample Proportions Formula: Given the information below, compute the confidence level associated with the Segmenting the College Student Marketing for Movie Attendance database. You will not need Excel to make this computation .

a. tolerable error is .05 or 5%

b. sample size = 500

c. standard deviation P = .5 and (1 – P) = .5

1. Confidence Level = ________________________%

2. How the results in this problem compare the confidence level in #2 of (3)?

____________________________________________________________

Chapter 15 – Data Processing and Fundamental Data Analysis

Exercise #1: Machine Cleaning Data

(1) Go to the Wiley website at www.wiley.com/college/mcdaniel and download the “Segmenting the College Student Market for Movie Attendance” Excel database. This database will have several errors for you to correct. Click on the “Value Labels” tab at the bottom of the Excel worksheet and notice the computer coding for each variable.

(2) Also from the Wiley website, download a copy of the “Segmenting the College Student Market for Movie Attendance” questionnaire. Notice the computer coding for each of the variables; it is the same as that in the “Value Labels” in the Excel worksheet. This information will be important in finding errors in the database.

(3) Now it’s time to check the database for errors. As you noticed in the “Value Labels”

worksheet, there are a limited number of possible answers to each question. For example, for Q1, “Did you attend at least one movie at a movie theatre in the past year?,” a respondent is only able to answer “Yes” or “No.” Therefore, the minimum value possible for that question is “0,” the numerical value label assigned to that response for the “No” response to for question; and the maximum value possible is “1,” the numerical label assigned to the “Yes” response for that question. In Excel, we can use formulas to find the minimum and maximum values among all 500 responses to any given question. So, if we find that the maximum value among all of the responses for Q1 is “2,” then we know that there was an error in the data entry process, because the only possible maximum value is “1,” meaning that the respondent answered “Yes” to the question. So,

Do This: In a blank cell, such as B503, enter this formula: =MIN(B2:B501)

You should get a calculated value of zero, which is what we expected.

Now, Do This: In the blank cell directly beneath the one you put the MIN formula in, say B504, enter this formula: =MAX(B2:B501)

You should get a calculated value of 1, theoretically, but you’ll see that you actually get a calculated value of 2, which means that someone messed up in entering the data. So, we need to find the error in the Q1 column.

Do This: Select the Q1 column (column B) to highlight it. Then click “Edit” on the toolbar at the top of the screen and select “find” (of for a shortcut, hold the “ctrl” button and hit “f,” which is the “find” command). The “find” window will pop up. Enter “2” in the box and hit enter. Excel will go to the location of the first 2 that it finds in the column for Q1.

Follow this procedure for finding errors in the columns for the rest of the variables. For example, for Q2, “Indicate how important you consider going to the movies at a movie theatre, relative to other leisure activities,” the minimum possible value is “1” for “Very Unimportant,” and the maximum possible value is “4” for “Very Important.” Since you have already entered the MIN and MAX formulas for Q1, all you really have to do is highlight cells B503 and B504, position the cursor so that you have the fill tool that looks like a cross, and drag the cursor all the way to the cells AE503 and AE504.

NOTE: If your spreadsheet contains values in rows 503 or 504 from completing the second exercise of Chapter 14 (computing standard deviations), you can delete them before filling in the cells or simply “drag” over them as shown in the subsequent screenshot.

When you analyze what you have just done, you see that:

· we have an error in the column for Q1 (as we already discussed)

· there are no errors in the column for Q2

· the MIN and MAX for Q3 don’t matter because it’s an open-ended question

· we have an error in the column for Q5f (the MAX value is 7, which is not a possible response)

· we have an error in the column for Q6

· the MIN and MAX for Q8a-d don’t matter because they are open-ended questions (however, we can check these for errors using another way, which we will do in a minute)

· we have an error in the column for Q9

· we have an error in the column for Q10

· there are no errors in the column for Q11

· we have an error in the column for Q12

· there are no errors in the column for Q13

· there are no errors in the column for Q14

(4) When you have determined which variables have input errors, summarize the errors using the template below as a guide.

Questionnaire Number

Variable containing error

Incorrect Value

Correct value

(5) Now, as we stated above, another possible source of errors is in question 8. Notice that in this question that the sum of the answers should be 100%. To check that these add up to 100% for each respondent, just enter the following formula into a blank cell for Questionnaire Number 1 (such as cell AF2): =SUM(V2:Y2) (there are also other ways to calculate the sum of certain cells using Excel; see if you can play around with the tools or if you can figure out varying but equivalent formulas will do the trick!)

Now all that you have to do to calculate the sum of Q8a-d for the rest of the database is to use the “fill” tool that you have most likely mastered by now! [Highlight cell AF2, position the cursor to get the “fill” tool, then “drag” the cursor all the way to the last Questionnaire entry.]

When you look over the calculated sums, you will notice that there is an error in the row for Questionnaire number 238, because the sum of Q8a-d is 110, when it should only be 100.

(6) Once you have completed summarizing the variables containing errors, position the cursor on each of the variable columns containing errors (one at a time) to highlight the column. Use the ctrl-f function to find the questionnaire numbers where the errors occurred. The best approach to handling the errors for this assignment is to treat them as missing values. To do so, simply delete the incorrect numbers and leave the corresponding cell blank. Be sure to resave your database after correcting it for errors.

(7) After machine cleaning your data, you will notice that the values for your MIN and MAX formulas for your corrected database are now in accordance with the correct range for the value labels.

Exercise #2: Analysis of Data with Frequency Distributions

Now that you have a clean database, let’s obtain frequencies for all of the variables! First, let’s create multiple workspaces: position your cursor on the topmost left corner of the worksheet and click it to highlight the entire worksheet.

Now hold “Ctrl” and hit “c” for the “copy” command. Click Sheet 2 on the bottom of the screen to open a blank worksheet, position the cursor in cell A1, and hold “Ctrl” and hit “v” for the “paste” command (you can also use the copy and paste icons on the toolbar at the top of the screen for these commands). You’ll see that you just created an exact replica of your original clean database. Do this again in about 5 or 6 other worksheets (you may not use them all), and name them each something unique whenever you start to use them by double-clicking on the word “Sheet 1” (etc.) to highlight it and replace the text.

To make it easier to work with your worksheets, you can “freeze” the panes that indicate the Questionnaire number and the Variables:

Highlight cell B2, click “Window” on the toolbar at the top of the screen, and select “Freeze Panes.” From now on, no matter where you are in the database, you will still be able to see the variable labels and the questionnaire numbers.

We will now use the COUNTIF functions to determine frequencies for each variable.

How many people answered “Yes” to Question 1?

In cell B503 enter the following formula: =COUNTIF(B2:B501,1)

Excel calculates that 450 respondents answered “Yes” to Q1

What percent of the respondents answered “Yes” to Q1?

In cell B504 enter the following formula: =B503/500 (remember 500 is the total number

of respondents)

Excel calculates that 90% of the respondents answered “Yes” to Q1

How many people answered “No” to Question 1?

In cell B506 enter the following formula: =COUNTIF(B2:B501,0)

Excel calculates that 50 respondents answered “No” to Q1

What percent of the respondents answered “No” to Q1?

In cell B507 enter the following formula: =B506/500

Excel calculates that 10% of the respondents answered “No” to Q1

Use this COUNTIF method to answer the following questions, but remember that for some questions, not all 500 respondents marked an answer. So, in cases where you are asked to find percentages, you will have to divide the COUNTIF calculation by the true number of respondents for that question.

Answer the following:

1. What percentage of all respondents attended at least 1 movie in the past year? ______%

2. What percentage of all respondents never buy food items at a movie? ______%

(Hint: {=COUNTIF(E2:E501,4)/500} )

3. Produce a table indicating the percentage of all respondents that consider each of the movie theatre items in Question 5 of the questionnaire very important. List the top 5 movie items in descending order (start with the movie item have the highest percentage of very important responses).

For Example:

Movie Item

Percentage of Respondents

Movie Item with the highest percentage

63.0%

Movie Item with the 2nd highest percentage, etc.

57.6%

4. What percentage of respondents consider the “newspaper” a very important

source of information about movies playing at movie theatres? _______%

5. What percentage of respondents consider the “Internet” a very unimportant source

of information about movies playing at movie theatres? _______%

6. By observing the distribution of responses for Q8a, Q8b, Q8c and Q8d, which is

the most popular purchase option for movie theatre tickets? _______________

7. Produce a table listing in descending order the percentage of respondents that consider each of the movie theatre information sources (Q7) very important.

For Example:

Movie Theatre Information Sources

Percentage of Respondents indicating Very Important

Internet

44.4%

Newspaper

25.8%

Exercise #3: Analysis of Data with Descriptive Statistics
The objective of this exercise is to analyze data using measures of central tendency and measures of dispersion. To analyze means and standard deviations, we will use the =AVERAGE() and =STDEV() functions. To analyze medians and modes, we will use the =MEDIAN() and =MODE() functions.

On the questionnaire, Question #5 utilizes a 4-point Itemized Rating scale (illustrated below). This scale is balanced and can be assumed to yield interval scale/metric data. Given the preceding, invoke SPSS to calculate the mean and standard deviation for all of the variables in Question 5 (Q5a-Q5i).

EXAMPLES:

Very

unimportant

Somewhat

unimportant

Somewhat

important

Very

important

1

2

3

4

Answer the following questions:

1. Using only the mean for each of the variables, which of the movie theatre items was considered “most important?” __________________

2. Using only the standard deviation for each of the variables, for which question

was there the greatest amount of agreement? __________________

( Hint: Least amount of dispersion regarding the response to the movie item)

3. Questions 4 & 6 utilize multiple choice questions which yield non-metric data,

but which is ordinal scale. The appropriate measures of central tendency for non-

metric data are the median and mode.

a. What is the median response for Question #4, concerning the amount a

person spends on food/drink items at a movie? ________________

Never buy food items at movies

(0)

Up to $7.49

(1)

$7.50 to $14.99

(2)

$15.00 or more

(3)

b. Concerning Question #6, the distance a person would drive to see a movie

on a “big screen,” what is the mode of that distribution of responses?

Zero

(0)

1 to 9 miles

(1)

11 to 24 miles

(2)

25 to 49 miles

(3)

50+ miles

(4)

4. In this question the objective will be to compare the results of median and mean responses for Q3.

a. Mean response: _________

b. Median response: _________

c. Standard Deviation: _________

d. Minimum response: _________

e. Maximum response: _________

5. When the responses to a question contain extreme values, the mean response can be lie in the upper or lower quartile of the response distribution. In such a case, the median value would be a better indicator of an average response than the mean value. Given the information you obtained from answering #4 above, is the Mean or Median a better representative of the “average” response to Q3?

__________________________________________________________________

__________________________________________________________________

__________________________________________________________________

Exercise #4: Analysis of Demographic Characteristics Using Charts
You will use the COUNTIF() method described above to obtain frequencies for the demographic questions (questions 11-14).

EXAMPLE:.

Answer the following questions.

1. Display the demographic data for each of the four demographic variables in tables.

a)

b) Fill in the other demographics by dragging the “fill” tool as in examples from previous exercises.

c)

2.

3. For each demographic variable, illustrate the table results using some type of graphic representation of the data (pie charts, line charts, or bar charts).

EXAMPLE:

(The data in the picture to the left would create a pie chart like the one on the right.)

Chapter 16 – Statistical Testing of Differences and Relationships

Exercise #1: Analyzing Data Using Crosstabulation Analysis

Go to the Wiley website at www.wiley.com/college/mcdaniel and download the worksheet for this exercise. In this exercise you will use Excel to construct a “Pivot Table” in order to perform a Chi-Square test for statistically significant correlations between specified variables.

In this exercise we are assessing whether or not persons who attend movies at movie theatres are demographically different from those who do not. We will use the following pairs of variables:

a. Q1 & Q11

b. Q1 & Q12

c. Q1 & Q13

d. Q1 & Q14

In the worksheet “Q1xQ11,” when you scroll down, you will see some tables. These tables contain analysis results for Q1 and Q11. The first table is the Pivot Table. The second table is a crosstabulation that references the pivot table and contains the actual counts of the variables in relation to one another.

The third, fourth, and fifth tables contain the row, column, and total percents relative to the variables.

The sixth table contains the expected count, which is the expected number in each cell if the two variables were linearly independent, meaning that there was no correlation between them.

The seventh table contains the elements for the Chi-Square value. Finally, the results of the Chi-Square test are at the bottom of the worksheet. The Chi-Square value is the sum of the numbers from the last table. The number of columns and rows has been entered already to determine the degrees of freedom. The CHIDIST function then determines the significance of the Chi-Square value.

You will now need to use these tables to analyze the other variable combinations above.

Do This:

Copy the worksheet three times (into three new worksheets like you did for a previous exercise).

Right Click on the Pivot Table.

Click "Pivot Table Wizard" on the pop-up menu and click "Finish".

Drag Q11 out to the Pivot Table into the Pivot Table Field List.

Drag Q12 from the Pivot Table Field List into the column header (from where you just removed Q11).

Your Pivot Table should have changed!

But Wait! Do you see "#REF!" in the tables underneath the Pivot Table???

If yes, then GOOD! Because that's what's supposed to happen! This happens because the Actual Count table is still looking for Q11! In the actual count table, change the references for each of the four conditions (e.g., Q1=0 and Q11=1, Q1=0 an Q11=1, etc.) by doing the following. In the cell for Q1=0 and Q11=0, type “=” and then select the coordinating cell in the pivot table. Repeat this process for the remaining cells. For Q12, since there are only two responses (i.e. 0 and 1) instead of four (i.e., 1,2,3,4), you will need to clear the contents for the two columns that are no longer in use. Also, for Q12 (only), you also have to change the column labels from a range of 1 to "blank" to a range of 0 to 1. For all of the remaining tables, you will need to adjust the column headings and CLEAR CONTENTS for the two columns not in use (i.e., labeled 3 and 4).

If you see the #### symbols in any table, this indicates that the calculated value won’t fit in the cell and that you have to resize it by clicking, holding, and dragging the line between column headings until your values fit.

Now do the crosstabulation analysis for the rest of the variable combinations that we are interested in for this exercise.

Answer questions 1-6 using only the sample data. Do not consider the results of the chi-square test.

1. What % of males do not attend movies at movie theatres? _________%

2. What % of all respondents are African-American and do not _________%

attend movies at movie theatres?

3. What % of respondents not attending movies at movie theatres _________%

are in the 19-20 age category?

4. Which classification group is most likely to attend movies _________

at movie theatres?

5. Which age category is least likely to attend movies at a _________

movie theatre?

6. Are Caucasians less likely to attend movies at movie theatres _________

than African-Americans?

For question 7, the objective is to determine statistically whether in the population from which the sample data was drawn, if there were demographic differences in persons who attend and do not attend movies at movie theatres. We do this by using the results of the chi-square test for independent samples .

7. Evaluate the chi-square statistic in each of your crosstab tables. Construct a table

to summarize the results.

For example:

Variables

Pearson Chi-Square

Degrees of Freedom

asymp

sig.

Explanation

Q1 (attend or not attend movies at movie theatres & Q12 (gender)

4.76

3

.189

We can be 90% confident that based on our sample results, males differ significantly from females in their tendency to attend or not attend movies at movie theatres.

Exercise #2: T/Z Test for Independent Samples.

Go to the Wiley website at www.wiley.com/college/mcdaniel and download the worksheet for this exercise. This exercise compares males and females (Q12) regarding the information sources they utilize to search for information about movies at movie theatres (Q7a-e). The data has been sorted by gender (Q12) for you. Males are in rows 2 through 228 and females are in rows 229 through 501.

Before we conduct the t-Tests, as you learned in class, the T/Z Test involves the tests for the Equality of Means and the Levene's Test for the Equality of Variance. The result of this latter test will indicate which functions in Excel to use.

For the Levene’s test, Do This:

Click “Tools” on the toolbar at the top of the screen, select “Data Analysis”, select “t-Test: Two-Sample Assuming Equal Variances”, and click “Ok”.

To run a test comparing the variances of male and female population for Q7a, in the box labeled Variable 1 range, enter: “B2:B228,” and in the box labeled Variable 2 range, enter: “B229:B501” (as shown below). Then click on “OK.”

The results will appear in a new sheet. We can see (below) that the variance test is not significant. Hence we can do a t-Test with equal variances assumed. If this test were significant, we would simply choose to do a t-Test with equal variances not assumed, and follow the exact same procedure we are about to detail.

For the t-Test, Do This:

Click “Tools” on the toolbar at the top of the screen, select “Data Analysis”, select “t-Test: Two-Sample Assuming Equal Variances”, and click “Ok”.

To run a t-test for Q7a, in the box labeled Variable 1 range, enter: “B2:B228,” and in the box labeled Variable 2 range, enter: “B229:B501” (as shown below). Then, enter a “0” in the box labeled hypothesized Mean Difference. Then click on “OK.”

In a new worksheet, you will see the results below. Please note: you can change the cells with scientific notation to a regular number by selecting the cell and selecting format from the menus.

Repeat this process and the test of equal variances for q7b-f.

When you do the t-test this way, you can look at the calculated t statistic and compare it to the critical t value. If the calculated t statistic is greater than the critical t value, then the test is significant, meaning males and females differ significantly concerning the sources by which they get their movie theatre information.

Answer the following questions:

From these results of our sample data, can we generalize our results to the population by saying that males differ from females regarding the importance they place on various information

sources to get information about movies at movie theatres by:

1. the newspaper (Q7a)?

2. the Internet (Q7b)?

4. phoning in to the movie theatre for information (Q7c)?

5. the television (Q7d)?

6. friends or family (Q7e)?

You may want to use the template below to summarize your T-test results. For example:

Variables

Mean Difference

Means

Prob of Sig diff

Interpretation of Results

Q12 (gender) & Q7a (newspaper)

.373

.000

99.9% confident that based on our sample results, males differ significantly from females concerning the importance they place on the newspaper as an information source about movies at movie theatres (means test).

Exercise #3: ANOVA Test for Independent Samples.

Go to the Wiley website www.wiley.com/college/mcdaniel and download the worksheet for this exercise.

This exercise compares the responses of freshmen, sophomores, juniors, seniors, and graduate students to test for significant differences in the importance placed on several movie theatre items. The ANOVA test produces a SUMMARY table based on sample data. Just like the t-test, if the ANOVA test is significant, then we can make inferences about the general population under study concerning which student classification places most importance on comfortable seats, for example. You’ll notice that the data in the worksheet has been grouped together (all of the responses that freshmen made for Q5a are grouped together, etc.) in a new worksheet called "Rearr Data.” This is necessary for Excel to run an ANOVA.

Now, Do This:

Click “Tools” on the toolbar at the top of the screen, select “Data Analysis”, select “ANOVA: Single Factor”, and click “Ok”.

Enter the following data range into the box: “A2:E155”, check the box by “Labels in First Row”, and hit “Ok”.

Look at the results:

Repeat the process and run ANOVA for the other variables.

Answer the following questions:

From our sample data, can we generalize our results to the population by saying that

there are significant differences across the classification of students by the importance

they place on the following movie theatre items?

1. video arcade at the movie theatre (Q5a)

2. soft drinks and food items Q5b)

3. plentiful restrooms (Q5c)

4. comfortable chairs (Q5d)

5. auditorium type seating (Q5e)

6. size of the movie theatre screens (Q5f)

7. quality of the sound system (Q5g)

8. number of screens at a movie theatre (Q5h)

9. clean restroom (Q5i)

10. Using only the descriptive statistics, which classification group (Q13) places the

least amount of importance on clean restrooms (Q5i)? ___________________

11. Using only the SUMMARY table, which classification group (Q13) places the

greatest amount of importance on quality of sound system (Q5i)?

___________________

Summarize the results of your ANOVA analysis using a table similar to the one below.

Variables

Degrees of Freedom

F-Value

Probability of Insignificance

Interpretation of Results

Q5a (importance of a video arcade) & Q13 (student classification)

4, 442

.766

.548

The p-value for this test is well above .05, hence, there is NO significant difference in importance of video arcades due to student level.

Ethnic Background

African-

American

50%

Caucasian

20%

Other

20%

Hispanic

10%