You are on page 1of 6

Excel_StudyCard_2008.

qxd 9/17/08 2:27 PM Page 1

Excel® 2007 with DDXL J E N N I F E R L E W I S P R I E S T L E Y, P H . D.


Scatterplot, Regression, and Residual Plots Correlation
Typically, scatterplots are generated as a first step to a regression FROM AN OPEN EXCEL 2007 WORKSHEET
exercise. Assuming this is the goal, use the following procedure to Use the following procedure to generate the correlation coefficient for two quantitative variables:
generate a scatterplot, a simple linear regression model, and the 1. Highlight the entire dataset, including the labels in the first row.
associated residual plot (the process to generate a simple stand-
Introduction to Excel 2007 with DDXL alone scatterplot is shown below): 2. Select Add-Ins>DDXL>Regression.
The layout for Excel 2007 is very different from the layout for earlier versions of 1. Highlight the entire dataset, including the labels in the 3. Under Function type, select Correlation. This will generate three boxes with arrows next to them. Under the Names and
Excel. However, with a little practice, students should find that the new layout is first row. Columns list, select a quantitative variable and click the arrow next to the x-Axis Quantitative Variable box. Then select a
very logical and easy to use. Once the DDXL add-in is installed, the Excel 2007 second quantitative variable and click the arrow next to the y-Axis Quantitative Variable box. Select OK.
2. Select Add-Ins>DDXL>Regression.
interface appears as shown to the right:
3. Under Function type, select Simple Regression. This will Finding the Area under the Normal Curve and Inverse Normality
The tabs across the top can be thought of as “tool boxes,” where the options with- generate three boxes with arrows next to them. Under the
in each tab represent the individual tools. The DDXL functionality can be found Names and Columns list, select a quantitative variable and FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to find the normal cumulative distribution and its inverse:
under the Add-Ins tab, as circled above: click the arrow next to the Response Variable box. Then 1. Select a quantitative variable to analyze. At the bottom of in front of the number of the cell reference. In the Cumulative
select a second quantitative variable and click the arrow next that variable column, generate the mean and the standard box, type True. Click OK.
to the Explanatory Variable box. Select OK. deviation. To generate the mean, enter the formula 5. The resulting value will be the cumulative probability (from
4. Typical results are shown to the right. =AVERAGE(A2:A41), where A2 through A41 is the range negative infinity) associated with the value in that row of the
5. To generate the residual plot, click on the Check the of the data. To generate the standard deviation for the sample, variable of interest, assuming a normal distribution. Copy
Residuals button. enter the formula =STDEV(A2:A41). this formula to the bottom of the column.
2. Once you have the mean and the standard deviation for the 6. To determine the inverse — the value of interest based upon a
variable at the bottom of the variable column, insert a new normal probability of occurrence — create a new column to
Descriptive Statistics blank column next to the variable of interest. To do this, click the right of the data of interest. Click on the fx button. In the
FROM AN OPEN EXCEL 2007 WORKSHEET Use the follow- 3. Under Function type, select Summary of One Variable. on the top of the column where you want a new column Search for a Function box, type Normal Distribution. From
This will generate a box entitled Quantitative Variable with Histogram and Boxplot inserted. Select Home>Insert. the Select a Function list, click on NORMINV. You will see
ing procedure to find the complete description of a quantitative
an arrow next to it. Under the Names and Columns list, 3. To find the associated area under the normal curve of values the following screen:
variable, including the mean, median, and standard deviation: FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to generate for a variable, go to the first empty cell in the new column.
select a quantitative variable and click the arrow next to the the histogram and boxplot for a quantitative variable:
1. Highlight the entire dataset, including the labels in the first row. Quantitative Variable box. Select OK. Click on the fx button. In the Search for a Function box,
1. Highlight the entire dataset, including the labels in the first row. type Normal Distribution. From the Select a Function list,
2. Select Add-Ins>DDXL>Summaries. You should see the 4. Typical results are shown below:
2. Select Add-Ins>DDXL>Charts and Plots. click on NORMDIST. You will see the following screen:
following screen:
3. To generate a histogram, under Function type, select Histogram. This will generate a box entitled Quantitative Variable with an
arrow next to it. Under the Names and Columns list, select a quantitative variable and click the arrow next to the Quantitative
Variable box. Select OK.
4. To generate a boxplot, under Function type, select Boxplot. This will generate a box entitled Quantitative Variable with an arrow
next to it. Under the Names and Columns list, select a quantitative variable and click the arrow next to the Quantitative Variable
box. Select OK.
5. To generate side-by-side boxplots to compare the spread of a quantitative variable across specific groups, select Boxplot by Groups
under Function type. Select a quantitative variable and click the arrow next to the Quantitative Variable box. Select a categorical
variable and click the arrow next to the Group Variable box. 7. In the Probability box, enter the cumulative probability that
you are interested in from the normal curve. In the Mean box,
click on the cell where you calculated the average for the vari-
Displaying Categorical Data (Frequency Tables) able of interest. Type a $ in front of the letter and then again in
FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to generate a frequency table: 4. In the X box, click on the first value in your variable list — front of the number, referencing the cell where the average was
1. Highlight the entire dataset, including the labels in the first row. from the ranges identified in the calculations above, it would be calculated. In the Standard_dev box, click on the cell where
A2. In the Mean box, click on the cell where you calculated the you calculated the STDEV. Again, type a $ in front of the letter
2. Select Add-Ins>DDXL>Tables. Under Function type, select Frequency Table. This will generate a box entitled Categorical Variable
average. Type a $ in front of the letter and then again in front of and in front of the number of the cell reference. Click OK.
with an arrow next to it. Under the Names and Columns list, select a categorical variable and click the arrow next to the box. Select OK.
the number, referencing the cell where the average was calcu- 8. The resulting value will be the cumulative observation asso-
lated. In the Standard_dev box, click on the cell where you ciated with the cumulative probability indicated, assuming a
Confidence Intervals for Means calculated the STDEV. Again, type a $ in front of the letter and normal distribution.
FROM AN OPEN EXCEL 2007 WORKSHEET Use the following
procedure to generate the 90%, 95%, or 99% confidence interval for a Assessing Normality
quantitative variable:
1. Highlight the entire dataset, including the labels in the first row. FROM AN OPEN EXCEL 2007 WORKSHEET
Use the following procedure to assess the normality (or not) of a quantitative variable of interest:
2. Select Add-Ins>DDXL>Confidence Intervals.
1. Highlight the entire dataset, including the labels in the first row.
ISBN-13: 978-0-321-59280-4 3. Under Function type, select 1 Var t Interval (Note: While the 1
ISBN-10: 0-321-59280-8
2. Select Add-Ins>DDXL>Charts and Plots.
Var z Interval is slightly more accurate, it requires the ability to
9 0 0 0 0 estimate the standard deviation of the population, which is rare.) 3. Under Function type, select Normal Probability Plot. This will generate two boxes with an arrow next to them. Under the Names
This will generate a box entitled Quantitative Variable with an arrow next to it. Under the Names and Columns list, select a quantita- and Columns list, select a quantitative variable and click the arrow next to the first box. Select OK.
tive variable and click the arrow next to the Quantitative Variable box. Select OK. You should see the screen above: 4. Note that a normally distributed variable will follow a 45-degree line in the scatterplot graphic (this graphic plots the actual values
4. Click on the confidence level that you need, and then click on the Compute Interval button. of the variable on the y-axis versus the z-scores on the x-axis).

1 9 780321 592804 2 3
Excel_StudyCard_2008.qxd 9/17/08 2:27 PM Page 1

Excel® 2007 with DDXL J E N N I F E R L E W I S P R I E S T L E Y, P H . D.


Scatterplot, Regression, and Residual Plots Correlation
Typically, scatterplots are generated as a first step to a regression FROM AN OPEN EXCEL 2007 WORKSHEET
exercise. Assuming this is the goal, use the following procedure to Use the following procedure to generate the correlation coefficient for two quantitative variables:
generate a scatterplot, a simple linear regression model, and the 1. Highlight the entire dataset, including the labels in the first row.
associated residual plot (the process to generate a simple stand-
Introduction to Excel 2007 with DDXL alone scatterplot is shown below): 2. Select Add-Ins>DDXL>Regression.
The layout for Excel 2007 is very different from the layout for earlier versions of 1. Highlight the entire dataset, including the labels in the 3. Under Function type, select Correlation. This will generate three boxes with arrows next to them. Under the Names and
Excel. However, with a little practice, students should find that the new layout is first row. Columns list, select a quantitative variable and click the arrow next to the x-Axis Quantitative Variable box. Then select a
very logical and easy to use. Once the DDXL add-in is installed, the Excel 2007 second quantitative variable and click the arrow next to the y-Axis Quantitative Variable box. Select OK.
2. Select Add-Ins>DDXL>Regression.
interface appears as shown to the right:
3. Under Function type, select Simple Regression. This will Finding the Area under the Normal Curve and Inverse Normality
The tabs across the top can be thought of as “tool boxes,” where the options with- generate three boxes with arrows next to them. Under the
in each tab represent the individual tools. The DDXL functionality can be found Names and Columns list, select a quantitative variable and FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to find the normal cumulative distribution and its inverse:
under the Add-Ins tab, as circled above: click the arrow next to the Response Variable box. Then 1. Select a quantitative variable to analyze. At the bottom of in front of the number of the cell reference. In the Cumulative
select a second quantitative variable and click the arrow next that variable column, generate the mean and the standard box, type True. Click OK.
to the Explanatory Variable box. Select OK. deviation. To generate the mean, enter the formula 5. The resulting value will be the cumulative probability (from
4. Typical results are shown to the right. =AVERAGE(A2:A41), where A2 through A41 is the range negative infinity) associated with the value in that row of the
5. To generate the residual plot, click on the Check the of the data. To generate the standard deviation for the sample, variable of interest, assuming a normal distribution. Copy
Residuals button. enter the formula =STDEV(A2:A41). this formula to the bottom of the column.
2. Once you have the mean and the standard deviation for the 6. To determine the inverse — the value of interest based upon a
variable at the bottom of the variable column, insert a new normal probability of occurrence — create a new column to
Descriptive Statistics blank column next to the variable of interest. To do this, click the right of the data of interest. Click on the fx button. In the
FROM AN OPEN EXCEL 2007 WORKSHEET Use the follow- 3. Under Function type, select Summary of One Variable. on the top of the column where you want a new column Search for a Function box, type Normal Distribution. From
This will generate a box entitled Quantitative Variable with Histogram and Boxplot inserted. Select Home>Insert. the Select a Function list, click on NORMINV. You will see
ing procedure to find the complete description of a quantitative
an arrow next to it. Under the Names and Columns list, 3. To find the associated area under the normal curve of values the following screen:
variable, including the mean, median, and standard deviation: FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to generate for a variable, go to the first empty cell in the new column.
select a quantitative variable and click the arrow next to the the histogram and boxplot for a quantitative variable:
1. Highlight the entire dataset, including the labels in the first row. Quantitative Variable box. Select OK. Click on the fx button. In the Search for a Function box,
1. Highlight the entire dataset, including the labels in the first row. type Normal Distribution. From the Select a Function list,
2. Select Add-Ins>DDXL>Summaries. You should see the 4. Typical results are shown below:
2. Select Add-Ins>DDXL>Charts and Plots. click on NORMDIST. You will see the following screen:
following screen:
3. To generate a histogram, under Function type, select Histogram. This will generate a box entitled Quantitative Variable with an
arrow next to it. Under the Names and Columns list, select a quantitative variable and click the arrow next to the Quantitative
Variable box. Select OK.
4. To generate a boxplot, under Function type, select Boxplot. This will generate a box entitled Quantitative Variable with an arrow
next to it. Under the Names and Columns list, select a quantitative variable and click the arrow next to the Quantitative Variable
box. Select OK.
5. To generate side-by-side boxplots to compare the spread of a quantitative variable across specific groups, select Boxplot by Groups
under Function type. Select a quantitative variable and click the arrow next to the Quantitative Variable box. Select a categorical
variable and click the arrow next to the Group Variable box. 7. In the Probability box, enter the cumulative probability that
you are interested in from the normal curve. In the Mean box,
click on the cell where you calculated the average for the vari-
Displaying Categorical Data (Frequency Tables) able of interest. Type a $ in front of the letter and then again in
FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to generate a frequency table: 4. In the X box, click on the first value in your variable list — front of the number, referencing the cell where the average was
1. Highlight the entire dataset, including the labels in the first row. from the ranges identified in the calculations above, it would be calculated. In the Standard_dev box, click on the cell where
A2. In the Mean box, click on the cell where you calculated the you calculated the STDEV. Again, type a $ in front of the letter
2. Select Add-Ins>DDXL>Tables. Under Function type, select Frequency Table. This will generate a box entitled Categorical Variable
average. Type a $ in front of the letter and then again in front of and in front of the number of the cell reference. Click OK.
with an arrow next to it. Under the Names and Columns list, select a categorical variable and click the arrow next to the box. Select OK.
the number, referencing the cell where the average was calcu- 8. The resulting value will be the cumulative observation asso-
lated. In the Standard_dev box, click on the cell where you ciated with the cumulative probability indicated, assuming a
Confidence Intervals for Means calculated the STDEV. Again, type a $ in front of the letter and normal distribution.
FROM AN OPEN EXCEL 2007 WORKSHEET Use the following
procedure to generate the 90%, 95%, or 99% confidence interval for a Assessing Normality
quantitative variable:
1. Highlight the entire dataset, including the labels in the first row. FROM AN OPEN EXCEL 2007 WORKSHEET
Use the following procedure to assess the normality (or not) of a quantitative variable of interest:
2. Select Add-Ins>DDXL>Confidence Intervals.
1. Highlight the entire dataset, including the labels in the first row.
ISBN-13: 978-0-321-59280-4 3. Under Function type, select 1 Var t Interval (Note: While the 1
ISBN-10: 0-321-59280-8
2. Select Add-Ins>DDXL>Charts and Plots.
Var z Interval is slightly more accurate, it requires the ability to
9 0 0 0 0 estimate the standard deviation of the population, which is rare.) 3. Under Function type, select Normal Probability Plot. This will generate two boxes with an arrow next to them. Under the Names
This will generate a box entitled Quantitative Variable with an arrow next to it. Under the Names and Columns list, select a quantita- and Columns list, select a quantitative variable and click the arrow next to the first box. Select OK.
tive variable and click the arrow next to the Quantitative Variable box. Select OK. You should see the screen above: 4. Note that a normally distributed variable will follow a 45-degree line in the scatterplot graphic (this graphic plots the actual values
4. Click on the confidence level that you need, and then click on the Compute Interval button. of the variable on the y-axis versus the z-scores on the x-axis).

1 9 780321 592804 2 3
Excel_StudyCard_2008.qxd 9/17/08 2:27 PM Page 1

Excel® 2007 with DDXL J E N N I F E R L E W I S P R I E S T L E Y, P H . D.


Scatterplot, Regression, and Residual Plots Correlation
Typically, scatterplots are generated as a first step to a regression FROM AN OPEN EXCEL 2007 WORKSHEET
exercise. Assuming this is the goal, use the following procedure to Use the following procedure to generate the correlation coefficient for two quantitative variables:
generate a scatterplot, a simple linear regression model, and the 1. Highlight the entire dataset, including the labels in the first row.
associated residual plot (the process to generate a simple stand-
Introduction to Excel 2007 with DDXL alone scatterplot is shown below): 2. Select Add-Ins>DDXL>Regression.
The layout for Excel 2007 is very different from the layout for earlier versions of 1. Highlight the entire dataset, including the labels in the 3. Under Function type, select Correlation. This will generate three boxes with arrows next to them. Under the Names and
Excel. However, with a little practice, students should find that the new layout is first row. Columns list, select a quantitative variable and click the arrow next to the x-Axis Quantitative Variable box. Then select a
very logical and easy to use. Once the DDXL add-in is installed, the Excel 2007 second quantitative variable and click the arrow next to the y-Axis Quantitative Variable box. Select OK.
2. Select Add-Ins>DDXL>Regression.
interface appears as shown to the right:
3. Under Function type, select Simple Regression. This will Finding the Area under the Normal Curve and Inverse Normality
The tabs across the top can be thought of as “tool boxes,” where the options with- generate three boxes with arrows next to them. Under the
in each tab represent the individual tools. The DDXL functionality can be found Names and Columns list, select a quantitative variable and FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to find the normal cumulative distribution and its inverse:
under the Add-Ins tab, as circled above: click the arrow next to the Response Variable box. Then 1. Select a quantitative variable to analyze. At the bottom of in front of the number of the cell reference. In the Cumulative
select a second quantitative variable and click the arrow next that variable column, generate the mean and the standard box, type True. Click OK.
to the Explanatory Variable box. Select OK. deviation. To generate the mean, enter the formula 5. The resulting value will be the cumulative probability (from
4. Typical results are shown to the right. =AVERAGE(A2:A41), where A2 through A41 is the range negative infinity) associated with the value in that row of the
5. To generate the residual plot, click on the Check the of the data. To generate the standard deviation for the sample, variable of interest, assuming a normal distribution. Copy
Residuals button. enter the formula =STDEV(A2:A41). this formula to the bottom of the column.
2. Once you have the mean and the standard deviation for the 6. To determine the inverse — the value of interest based upon a
variable at the bottom of the variable column, insert a new normal probability of occurrence — create a new column to
Descriptive Statistics blank column next to the variable of interest. To do this, click the right of the data of interest. Click on the fx button. In the
FROM AN OPEN EXCEL 2007 WORKSHEET Use the follow- 3. Under Function type, select Summary of One Variable. on the top of the column where you want a new column Search for a Function box, type Normal Distribution. From
This will generate a box entitled Quantitative Variable with Histogram and Boxplot inserted. Select Home>Insert. the Select a Function list, click on NORMINV. You will see
ing procedure to find the complete description of a quantitative
an arrow next to it. Under the Names and Columns list, 3. To find the associated area under the normal curve of values the following screen:
variable, including the mean, median, and standard deviation: FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to generate for a variable, go to the first empty cell in the new column.
select a quantitative variable and click the arrow next to the the histogram and boxplot for a quantitative variable:
1. Highlight the entire dataset, including the labels in the first row. Quantitative Variable box. Select OK. Click on the fx button. In the Search for a Function box,
1. Highlight the entire dataset, including the labels in the first row. type Normal Distribution. From the Select a Function list,
2. Select Add-Ins>DDXL>Summaries. You should see the 4. Typical results are shown below:
2. Select Add-Ins>DDXL>Charts and Plots. click on NORMDIST. You will see the following screen:
following screen:
3. To generate a histogram, under Function type, select Histogram. This will generate a box entitled Quantitative Variable with an
arrow next to it. Under the Names and Columns list, select a quantitative variable and click the arrow next to the Quantitative
Variable box. Select OK.
4. To generate a boxplot, under Function type, select Boxplot. This will generate a box entitled Quantitative Variable with an arrow
next to it. Under the Names and Columns list, select a quantitative variable and click the arrow next to the Quantitative Variable
box. Select OK.
5. To generate side-by-side boxplots to compare the spread of a quantitative variable across specific groups, select Boxplot by Groups
under Function type. Select a quantitative variable and click the arrow next to the Quantitative Variable box. Select a categorical
variable and click the arrow next to the Group Variable box. 7. In the Probability box, enter the cumulative probability that
you are interested in from the normal curve. In the Mean box,
click on the cell where you calculated the average for the vari-
Displaying Categorical Data (Frequency Tables) able of interest. Type a $ in front of the letter and then again in
FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to generate a frequency table: 4. In the X box, click on the first value in your variable list — front of the number, referencing the cell where the average was
1. Highlight the entire dataset, including the labels in the first row. from the ranges identified in the calculations above, it would be calculated. In the Standard_dev box, click on the cell where
A2. In the Mean box, click on the cell where you calculated the you calculated the STDEV. Again, type a $ in front of the letter
2. Select Add-Ins>DDXL>Tables. Under Function type, select Frequency Table. This will generate a box entitled Categorical Variable
average. Type a $ in front of the letter and then again in front of and in front of the number of the cell reference. Click OK.
with an arrow next to it. Under the Names and Columns list, select a categorical variable and click the arrow next to the box. Select OK.
the number, referencing the cell where the average was calcu- 8. The resulting value will be the cumulative observation asso-
lated. In the Standard_dev box, click on the cell where you ciated with the cumulative probability indicated, assuming a
Confidence Intervals for Means calculated the STDEV. Again, type a $ in front of the letter and normal distribution.
FROM AN OPEN EXCEL 2007 WORKSHEET Use the following
procedure to generate the 90%, 95%, or 99% confidence interval for a Assessing Normality
quantitative variable:
1. Highlight the entire dataset, including the labels in the first row. FROM AN OPEN EXCEL 2007 WORKSHEET
Use the following procedure to assess the normality (or not) of a quantitative variable of interest:
2. Select Add-Ins>DDXL>Confidence Intervals.
1. Highlight the entire dataset, including the labels in the first row.
ISBN-13: 978-0-321-59280-4 3. Under Function type, select 1 Var t Interval (Note: While the 1
ISBN-10: 0-321-59280-8
2. Select Add-Ins>DDXL>Charts and Plots.
Var z Interval is slightly more accurate, it requires the ability to
9 0 0 0 0 estimate the standard deviation of the population, which is rare.) 3. Under Function type, select Normal Probability Plot. This will generate two boxes with an arrow next to them. Under the Names
This will generate a box entitled Quantitative Variable with an arrow next to it. Under the Names and Columns list, select a quantita- and Columns list, select a quantitative variable and click the arrow next to the first box. Select OK.
tive variable and click the arrow next to the Quantitative Variable box. Select OK. You should see the screen above: 4. Note that a normally distributed variable will follow a 45-degree line in the scatterplot graphic (this graphic plots the actual values
4. Click on the confidence level that you need, and then click on the Compute Interval button. of the variable on the y-axis versus the z-scores on the x-axis).

1 9 780321 592804 2 3
Excel_StudyCard_2008.qxd 9/17/08 2:27 PM Page 2

Excel 2007 with DDXL


Generating Random Data and Sampling Hypothesis Test for Proportions (cont.) Hypothesis Test for Paired Differences
FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to generate random numbers and a simple random sample: which the proportion interval needs to be generated. The 7. In step 4, you are asked to determine if the test should be a FROM AN OPEN EXCEL 2007 WORKSHEET Use the following
1. Create a new column on the right of your dataset entitled values will scroll in alphabetical order. one-tail (less than or greater than) or a two-tail test. procedure to determine if a statistical difference exists between
Random. 5. In step 2, you are asked to designate the proportion value for 8. In step 5, click the Compute button. paired samples:
2. Inside the first open cell (row 2), type the following which the sample proportion will be tested against. 1. Highlight the entire dataset, including the labels in the first row.
function: =RAND(). 6. In step 3, you must assign the significance level 2. Select Add-Ins>DDXL>Hypothesis Tests.
3. After you click Enter, a random number between 0 and 1 (1— the confidence level).
3. Under Function type, select Paired t Test. This will generate
should have been generated. Note that this function RAND three boxes with arrows next to them. Under the Names and
generates random numbers that follow a uniform distribution Hypothesis Test for the Difference between Proportions Columns list, select a quantitative variable (e.g., a prescore)
between 0 and 1. Once your random number has been generated, and click the arrow next to the first box. Select a second
from the Home tab, click Copy. The cell should have what 6. Select the Sort by drop-down box and select RANDOM. FROM AN OPEN EXCEL 2007 WORKSHEET Use the follow-
ing procedure to generate a hypothesis test for the difference quantitative variable (e.g., a postscore) and click the arrow
appears to be little chasing lights around it. Now highlight the Select OK. next to the second box. Select OK. This will generate the
remainder of the column to the end of the data and click Paste. between two proportions:
7. Create a new spreadsheet entitled Sample. To do this, select screen shown to the right:
4. You may have noticed that the first value in row 2 changed. Home and from the Insert tab select Insert Sheet. Double 1. Highlight the entire dataset, including the labels in the first row.
2. Select Add-Ins>DDXL>Hypothesis Tests. 4. In step 1, click on the Set button and identify the
RAND is a volatile function in Excel, meaning that the result click on the Sheet 1 tab at the bottom of the worksheet and
hypothesized difference (typically, but not always, 0).
will change whenever a change is made in the spreadsheet. type Sample. 3. Under Function type, select 2 Var Prop Test. This will gen-
To resolve this, highlight the entire Random column, and 8. Return to the original datasheet and highlight the first N rows erate two boxes with arrows next to them. Under the Names 5. In step 2, select the appropriate significance level.
select Copy and then under the Paste options, select Paste (the number of observations you require for your sample). and Columns list, select a categorical variable and click the 6. In step 3, click on the Alternative Hypothesis option. generate three boxes with arrows next to them. Under the
Values. Now the values should be fixed. Make sure that the labels in the first row are included. Select arrow next to the first box. Select a second categorical vari- Names and Columns list, assign the same quantitative
7. In step 4, click on the Compute button.
5. Sort the data using the Random column. To do this, highlight Home>Copy. Go to the new Sample worksheet and click able and click the arrow next to the second box. Select OK. variables as done previously.
This will generate the screen shown to the right: 8. To generate the confidence interval for the paired differences,
the entire dataset. Select Data>Sort. The screen shown inside cell A1 in the upper-left corner and select 10. Select the appropriate confidence level and click on the
Add-Ins>DDXL>Confidence Intervals.
above to the right should appear: Home>Paste. 4. In step 1, click on the button Set Success 1 —one of the Compute button.
two proportion values that you want to test from the first 9. Under Function type, select Paired t Interval. This will
Confidence Intervals for Proportions variable—until the correct value is assigned. Click on the
Set Success 2 button —one of the two proportion values that
FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to generate a confidence interval for a proportion: you want to test from the second variable—until the correct
1. Highlight the entire dataset, including the labels in the first row.
value is assigned. 10. Under Function type, select 2 Var Prop Interval. This will Hypothesis Tests for Difference of Means in Two Independent Samples
5. In step 2, set the hypothesized difference between the propor- generate two boxes with arrows next to them. Under the
2. Select Add-Ins>DDXL>Confidence Intervals. Names and Columns list, select the same categorical vari- FROM AN OPEN EXCEL 2007 WORKSHEET Use the following
tions (this is typically, although not always, 0).
ables as done previously. Select OK. procedure to determine if a statistical difference exists between two
3. Under Function type, select 1 Var Prop Interval. This will 6. In step 3, select the appropriate significance level. independent sample means:
generate a box entitled Proportions Variable with an arrow 7. In step 4, select the appropriate test — one-tailed less than, 11. In step 1, set the same success values as done previously.
next to it. Under the Names and Columns list, select a cate- 1. Highlight the entire dataset, including the labels in the first row.
two tailed, or one-tailed greater than. 12. In step 2, set the desired confidence level.
gorical variable and click the arrow next to the box. Select 2. Select Add-Ins>DDXL>Hypothesis Tests.
OK. This will generate the screen shown to the right: 8. In step 5, select Compute. 13. In step 3, select Compute.
9. To generate the confidence interval of the difference, select 3. Under Function type, select 2 Var t Test. This will generate
4. In step 1, you are asked to designate the success condition— Add-Ins>DDXL>Confidence Intervals. two boxes with arrows next to them. Under the Names and
one of the two proportion values that you want to build the Columns list, select a quantitative variable and click the
interval around. Here, click on the Set Success button until arrow next to the first box. Select a second quantitative vari-
you see the value for which the proportion interval needs to Hypothesis Tests for One Sample Mean able and click the arrow next to the second box. Select OK.
be generated. The values will scroll in alphabetical order. FROM AN OPEN EXCEL 2007 WORKSHEET Use the following This will generate the screen shown to the right:
5. In step 2, you are asked to assign a 90%, 95%, or 99% procedure to determine if a statistical difference exists between a 4. In step 1, identify if the test is pooled or two sample.
confidence level. single sample mean and a specified value: 5. In step 2, specify the hypothesized difference (this is
6. In step 3, click the Compute Interval button. 1. Highlight the entire dataset, including the labels in the first row. typically, but not always, 0).
2. Select Add-Ins>DDXL>Hypothesis Tests. 6. In step 3, select the required significance level.
Hypothesis Test for Proportions 3. Under Function type, select 1 Var t Test (Note: While the 1 7. In step 4, select the appropriate test — one-tailed less than, 10. Under Function type, select 2 Var t Interval. This will gen-
Var z Test is slightly more accurate, it requires the ability to two-tailed, or one-tailed greater than. erate two boxes with arrows next to them. Under the Names
FROM AN OPEN EXCEL 2007 WORKSHEET Use the estimate the standard deviation of the population, which is and Columns list, select the same quantitative variables as
following procedure to generate a confidence interval for a rare.) This will generate one box with an arrow next to it. 8. In step 5, select Compute. done previously. Select OK.
proportion: Under the Names and Columns list, select a quantitative 9. To generate the confidence interval of the difference, select 11. In step 1, identify if the interval is pooled or two sample as
1. Highlight the entire dataset, including the labels in the first row. variable and click the arrow next to the box. Select OK. This Add-Ins>DDXL>Confidence Intervals. shown above.
2. Select Add-Ins>DDXL>Hypothesis Tests. will generate the screen shown to the right: 12. In step 2, set the desired significance level.
3. Under Function type, select 1 Var Prop Test. This will gen- 4. In step 1, click on the Set button and identify the specified
13. In step 3, select Compute.
erate a box entitled Proportions Variable with an arrow next value of interest that you want to test the sample mean
to it. Under the Names and Columns list, select a categorical against. 6. In step 3, click on the Alternative Hypothesis option.
variable and click the arrow next to the box. Select OK. This 5. In step 2, select the appropriate significance level. 7. In step 4, click on the Compute button.
will generate the screen shown to the right:
4. In step 1, you are asked to designate the success condition —
one of the two proportion values that you want to test. Here,
click on the Set Success button until you see the value for

4 5 6
Excel_StudyCard_2008.qxd 9/17/08 2:27 PM Page 2

Excel 2007 with DDXL


Generating Random Data and Sampling Hypothesis Test for Proportions (cont.) Hypothesis Test for Paired Differences
FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to generate random numbers and a simple random sample: which the proportion interval needs to be generated. The 7. In step 4, you are asked to determine if the test should be a FROM AN OPEN EXCEL 2007 WORKSHEET Use the following
1. Create a new column on the right of your dataset entitled values will scroll in alphabetical order. one-tail (less than or greater than) or a two-tail test. procedure to determine if a statistical difference exists between
Random. 5. In step 2, you are asked to designate the proportion value for 8. In step 5, click the Compute button. paired samples:
2. Inside the first open cell (row 2), type the following which the sample proportion will be tested against. 1. Highlight the entire dataset, including the labels in the first row.
function: =RAND(). 6. In step 3, you must assign the significance level 2. Select Add-Ins>DDXL>Hypothesis Tests.
3. After you click Enter, a random number between 0 and 1 (1— the confidence level).
3. Under Function type, select Paired t Test. This will generate
should have been generated. Note that this function RAND three boxes with arrows next to them. Under the Names and
generates random numbers that follow a uniform distribution Hypothesis Test for the Difference between Proportions Columns list, select a quantitative variable (e.g., a prescore)
between 0 and 1. Once your random number has been generated, and click the arrow next to the first box. Select a second
from the Home tab, click Copy. The cell should have what 6. Select the Sort by drop-down box and select RANDOM. FROM AN OPEN EXCEL 2007 WORKSHEET Use the follow-
ing procedure to generate a hypothesis test for the difference quantitative variable (e.g., a postscore) and click the arrow
appears to be little chasing lights around it. Now highlight the Select OK. next to the second box. Select OK. This will generate the
remainder of the column to the end of the data and click Paste. between two proportions:
7. Create a new spreadsheet entitled Sample. To do this, select screen shown to the right:
4. You may have noticed that the first value in row 2 changed. Home and from the Insert tab select Insert Sheet. Double 1. Highlight the entire dataset, including the labels in the first row.
2. Select Add-Ins>DDXL>Hypothesis Tests. 4. In step 1, click on the Set button and identify the
RAND is a volatile function in Excel, meaning that the result click on the Sheet 1 tab at the bottom of the worksheet and
hypothesized difference (typically, but not always, 0).
will change whenever a change is made in the spreadsheet. type Sample. 3. Under Function type, select 2 Var Prop Test. This will gen-
To resolve this, highlight the entire Random column, and 8. Return to the original datasheet and highlight the first N rows erate two boxes with arrows next to them. Under the Names 5. In step 2, select the appropriate significance level.
select Copy and then under the Paste options, select Paste (the number of observations you require for your sample). and Columns list, select a categorical variable and click the 6. In step 3, click on the Alternative Hypothesis option. generate three boxes with arrows next to them. Under the
Values. Now the values should be fixed. Make sure that the labels in the first row are included. Select arrow next to the first box. Select a second categorical vari- Names and Columns list, assign the same quantitative
7. In step 4, click on the Compute button.
5. Sort the data using the Random column. To do this, highlight Home>Copy. Go to the new Sample worksheet and click able and click the arrow next to the second box. Select OK. variables as done previously.
This will generate the screen shown to the right: 8. To generate the confidence interval for the paired differences,
the entire dataset. Select Data>Sort. The screen shown inside cell A1 in the upper-left corner and select 10. Select the appropriate confidence level and click on the
Add-Ins>DDXL>Confidence Intervals.
above to the right should appear: Home>Paste. 4. In step 1, click on the button Set Success 1 —one of the Compute button.
two proportion values that you want to test from the first 9. Under Function type, select Paired t Interval. This will
Confidence Intervals for Proportions variable—until the correct value is assigned. Click on the
Set Success 2 button —one of the two proportion values that
FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to generate a confidence interval for a proportion: you want to test from the second variable—until the correct
1. Highlight the entire dataset, including the labels in the first row.
value is assigned. 10. Under Function type, select 2 Var Prop Interval. This will Hypothesis Tests for Difference of Means in Two Independent Samples
5. In step 2, set the hypothesized difference between the propor- generate two boxes with arrows next to them. Under the
2. Select Add-Ins>DDXL>Confidence Intervals. Names and Columns list, select the same categorical vari- FROM AN OPEN EXCEL 2007 WORKSHEET Use the following
tions (this is typically, although not always, 0).
ables as done previously. Select OK. procedure to determine if a statistical difference exists between two
3. Under Function type, select 1 Var Prop Interval. This will 6. In step 3, select the appropriate significance level. independent sample means:
generate a box entitled Proportions Variable with an arrow 7. In step 4, select the appropriate test — one-tailed less than, 11. In step 1, set the same success values as done previously.
next to it. Under the Names and Columns list, select a cate- 1. Highlight the entire dataset, including the labels in the first row.
two tailed, or one-tailed greater than. 12. In step 2, set the desired confidence level.
gorical variable and click the arrow next to the box. Select 2. Select Add-Ins>DDXL>Hypothesis Tests.
OK. This will generate the screen shown to the right: 8. In step 5, select Compute. 13. In step 3, select Compute.
9. To generate the confidence interval of the difference, select 3. Under Function type, select 2 Var t Test. This will generate
4. In step 1, you are asked to designate the success condition— Add-Ins>DDXL>Confidence Intervals. two boxes with arrows next to them. Under the Names and
one of the two proportion values that you want to build the Columns list, select a quantitative variable and click the
interval around. Here, click on the Set Success button until arrow next to the first box. Select a second quantitative vari-
you see the value for which the proportion interval needs to Hypothesis Tests for One Sample Mean able and click the arrow next to the second box. Select OK.
be generated. The values will scroll in alphabetical order. FROM AN OPEN EXCEL 2007 WORKSHEET Use the following This will generate the screen shown to the right:
5. In step 2, you are asked to assign a 90%, 95%, or 99% procedure to determine if a statistical difference exists between a 4. In step 1, identify if the test is pooled or two sample.
confidence level. single sample mean and a specified value: 5. In step 2, specify the hypothesized difference (this is
6. In step 3, click the Compute Interval button. 1. Highlight the entire dataset, including the labels in the first row. typically, but not always, 0).
2. Select Add-Ins>DDXL>Hypothesis Tests. 6. In step 3, select the required significance level.
Hypothesis Test for Proportions 3. Under Function type, select 1 Var t Test (Note: While the 1 7. In step 4, select the appropriate test — one-tailed less than, 10. Under Function type, select 2 Var t Interval. This will gen-
Var z Test is slightly more accurate, it requires the ability to two-tailed, or one-tailed greater than. erate two boxes with arrows next to them. Under the Names
FROM AN OPEN EXCEL 2007 WORKSHEET Use the estimate the standard deviation of the population, which is and Columns list, select the same quantitative variables as
following procedure to generate a confidence interval for a rare.) This will generate one box with an arrow next to it. 8. In step 5, select Compute. done previously. Select OK.
proportion: Under the Names and Columns list, select a quantitative 9. To generate the confidence interval of the difference, select 11. In step 1, identify if the interval is pooled or two sample as
1. Highlight the entire dataset, including the labels in the first row. variable and click the arrow next to the box. Select OK. This Add-Ins>DDXL>Confidence Intervals. shown above.
2. Select Add-Ins>DDXL>Hypothesis Tests. will generate the screen shown to the right: 12. In step 2, set the desired significance level.
3. Under Function type, select 1 Var Prop Test. This will gen- 4. In step 1, click on the Set button and identify the specified
13. In step 3, select Compute.
erate a box entitled Proportions Variable with an arrow next value of interest that you want to test the sample mean
to it. Under the Names and Columns list, select a categorical against. 6. In step 3, click on the Alternative Hypothesis option.
variable and click the arrow next to the box. Select OK. This 5. In step 2, select the appropriate significance level. 7. In step 4, click on the Compute button.
will generate the screen shown to the right:
4. In step 1, you are asked to designate the success condition —
one of the two proportion values that you want to test. Here,
click on the Set Success button until you see the value for

4 5 6
Excel_StudyCard_2008.qxd 9/17/08 2:27 PM Page 2

Excel 2007 with DDXL


Generating Random Data and Sampling Hypothesis Test for Proportions (cont.) Hypothesis Test for Paired Differences
FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to generate random numbers and a simple random sample: which the proportion interval needs to be generated. The 7. In step 4, you are asked to determine if the test should be a FROM AN OPEN EXCEL 2007 WORKSHEET Use the following
1. Create a new column on the right of your dataset entitled values will scroll in alphabetical order. one-tail (less than or greater than) or a two-tail test. procedure to determine if a statistical difference exists between
Random. 5. In step 2, you are asked to designate the proportion value for 8. In step 5, click the Compute button. paired samples:
2. Inside the first open cell (row 2), type the following which the sample proportion will be tested against. 1. Highlight the entire dataset, including the labels in the first row.
function: =RAND(). 6. In step 3, you must assign the significance level 2. Select Add-Ins>DDXL>Hypothesis Tests.
3. After you click Enter, a random number between 0 and 1 (1— the confidence level).
3. Under Function type, select Paired t Test. This will generate
should have been generated. Note that this function RAND three boxes with arrows next to them. Under the Names and
generates random numbers that follow a uniform distribution Hypothesis Test for the Difference between Proportions Columns list, select a quantitative variable (e.g., a prescore)
between 0 and 1. Once your random number has been generated, and click the arrow next to the first box. Select a second
from the Home tab, click Copy. The cell should have what 6. Select the Sort by drop-down box and select RANDOM. FROM AN OPEN EXCEL 2007 WORKSHEET Use the follow-
ing procedure to generate a hypothesis test for the difference quantitative variable (e.g., a postscore) and click the arrow
appears to be little chasing lights around it. Now highlight the Select OK. next to the second box. Select OK. This will generate the
remainder of the column to the end of the data and click Paste. between two proportions:
7. Create a new spreadsheet entitled Sample. To do this, select screen shown to the right:
4. You may have noticed that the first value in row 2 changed. Home and from the Insert tab select Insert Sheet. Double 1. Highlight the entire dataset, including the labels in the first row.
2. Select Add-Ins>DDXL>Hypothesis Tests. 4. In step 1, click on the Set button and identify the
RAND is a volatile function in Excel, meaning that the result click on the Sheet 1 tab at the bottom of the worksheet and
hypothesized difference (typically, but not always, 0).
will change whenever a change is made in the spreadsheet. type Sample. 3. Under Function type, select 2 Var Prop Test. This will gen-
To resolve this, highlight the entire Random column, and 8. Return to the original datasheet and highlight the first N rows erate two boxes with arrows next to them. Under the Names 5. In step 2, select the appropriate significance level.
select Copy and then under the Paste options, select Paste (the number of observations you require for your sample). and Columns list, select a categorical variable and click the 6. In step 3, click on the Alternative Hypothesis option. generate three boxes with arrows next to them. Under the
Values. Now the values should be fixed. Make sure that the labels in the first row are included. Select arrow next to the first box. Select a second categorical vari- Names and Columns list, assign the same quantitative
7. In step 4, click on the Compute button.
5. Sort the data using the Random column. To do this, highlight Home>Copy. Go to the new Sample worksheet and click able and click the arrow next to the second box. Select OK. variables as done previously.
This will generate the screen shown to the right: 8. To generate the confidence interval for the paired differences,
the entire dataset. Select Data>Sort. The screen shown inside cell A1 in the upper-left corner and select 10. Select the appropriate confidence level and click on the
Add-Ins>DDXL>Confidence Intervals.
above to the right should appear: Home>Paste. 4. In step 1, click on the button Set Success 1 —one of the Compute button.
two proportion values that you want to test from the first 9. Under Function type, select Paired t Interval. This will
Confidence Intervals for Proportions variable—until the correct value is assigned. Click on the
Set Success 2 button —one of the two proportion values that
FROM AN OPEN EXCEL 2007 WORKSHEET Use the following procedure to generate a confidence interval for a proportion: you want to test from the second variable—until the correct
1. Highlight the entire dataset, including the labels in the first row.
value is assigned. 10. Under Function type, select 2 Var Prop Interval. This will Hypothesis Tests for Difference of Means in Two Independent Samples
5. In step 2, set the hypothesized difference between the propor- generate two boxes with arrows next to them. Under the
2. Select Add-Ins>DDXL>Confidence Intervals. Names and Columns list, select the same categorical vari- FROM AN OPEN EXCEL 2007 WORKSHEET Use the following
tions (this is typically, although not always, 0).
ables as done previously. Select OK. procedure to determine if a statistical difference exists between two
3. Under Function type, select 1 Var Prop Interval. This will 6. In step 3, select the appropriate significance level. independent sample means:
generate a box entitled Proportions Variable with an arrow 7. In step 4, select the appropriate test — one-tailed less than, 11. In step 1, set the same success values as done previously.
next to it. Under the Names and Columns list, select a cate- 1. Highlight the entire dataset, including the labels in the first row.
two tailed, or one-tailed greater than. 12. In step 2, set the desired confidence level.
gorical variable and click the arrow next to the box. Select 2. Select Add-Ins>DDXL>Hypothesis Tests.
OK. This will generate the screen shown to the right: 8. In step 5, select Compute. 13. In step 3, select Compute.
9. To generate the confidence interval of the difference, select 3. Under Function type, select 2 Var t Test. This will generate
4. In step 1, you are asked to designate the success condition— Add-Ins>DDXL>Confidence Intervals. two boxes with arrows next to them. Under the Names and
one of the two proportion values that you want to build the Columns list, select a quantitative variable and click the
interval around. Here, click on the Set Success button until arrow next to the first box. Select a second quantitative vari-
you see the value for which the proportion interval needs to Hypothesis Tests for One Sample Mean able and click the arrow next to the second box. Select OK.
be generated. The values will scroll in alphabetical order. FROM AN OPEN EXCEL 2007 WORKSHEET Use the following This will generate the screen shown to the right:
5. In step 2, you are asked to assign a 90%, 95%, or 99% procedure to determine if a statistical difference exists between a 4. In step 1, identify if the test is pooled or two sample.
confidence level. single sample mean and a specified value: 5. In step 2, specify the hypothesized difference (this is
6. In step 3, click the Compute Interval button. 1. Highlight the entire dataset, including the labels in the first row. typically, but not always, 0).
2. Select Add-Ins>DDXL>Hypothesis Tests. 6. In step 3, select the required significance level.
Hypothesis Test for Proportions 3. Under Function type, select 1 Var t Test (Note: While the 1 7. In step 4, select the appropriate test — one-tailed less than, 10. Under Function type, select 2 Var t Interval. This will gen-
Var z Test is slightly more accurate, it requires the ability to two-tailed, or one-tailed greater than. erate two boxes with arrows next to them. Under the Names
FROM AN OPEN EXCEL 2007 WORKSHEET Use the estimate the standard deviation of the population, which is and Columns list, select the same quantitative variables as
following procedure to generate a confidence interval for a rare.) This will generate one box with an arrow next to it. 8. In step 5, select Compute. done previously. Select OK.
proportion: Under the Names and Columns list, select a quantitative 9. To generate the confidence interval of the difference, select 11. In step 1, identify if the interval is pooled or two sample as
1. Highlight the entire dataset, including the labels in the first row. variable and click the arrow next to the box. Select OK. This Add-Ins>DDXL>Confidence Intervals. shown above.
2. Select Add-Ins>DDXL>Hypothesis Tests. will generate the screen shown to the right: 12. In step 2, set the desired significance level.
3. Under Function type, select 1 Var Prop Test. This will gen- 4. In step 1, click on the Set button and identify the specified
13. In step 3, select Compute.
erate a box entitled Proportions Variable with an arrow next value of interest that you want to test the sample mean
to it. Under the Names and Columns list, select a categorical against. 6. In step 3, click on the Alternative Hypothesis option.
variable and click the arrow next to the box. Select OK. This 5. In step 2, select the appropriate significance level. 7. In step 4, click on the Compute button.
will generate the screen shown to the right:
4. In step 1, you are asked to designate the success condition —
one of the two proportion values that you want to test. Here,
click on the Set Success button until you see the value for

4 5 6

You might also like