Up

                            t Tests with Excel

This exercise asks you to look at the attitude data set used in Chapter 8 and to analyze some data employing the t test for independent samples. You may complete the assignment by using either SPSS or Excel.  If you wish to use SPSS, do not use this page, but go to the page t Tests with SPSS.

The basic background of the study was as follows.  A researcher was interested in whether non sequitur fallacies were persuasive.   Argumentation theorists will tell you that such fallacies should not influence people, but they have serious doubts about responses of people who have never studied fallacies.  Thus the directional hypothesis in this study is that arguments using non sequitur appeals will produce lower levels of attitude change than arguments not using non sequitur appeals. In this study a couple of arguments were composed and the experimental treatment consisted of a message which included non sequitur reasoning focusing on largely irrelevant information.  In the control condition, the non sequitur reasoning was deleted, leaving a fairly sound argument.   These two conditions are the levels of the independent variable.  The variable that identified this fact is "falexper" (1= with the fallacy presented to subjects; 2=without the fallacy presented to subjects).  The dependent variable is "attitude" toward the topic (as created in chapter 8), which is the sum of the scores of the scales "wise-foolish;" "good-bad;" "positive-negative;" and "beneficial-harmful".

The task in this activity is to determine if there is a significant difference between the two conditions in this experiment. Complete a t test with alpha risk at .05, one-tailed.

If you have not already done so, go to the Data Sets location and download the file "ATTITUD1.SAV" under the "Excel Files" category.

Start your Excel program.  From the "File" menu Open the data file "ATTITUD1.SAV" from the location where the you saved it.

bulletWith Excel, the dependent score conditions must be in separate columns. Thus, we must separate the attitude data into two columns, one for each group.  In column Y row 1 enter "fallwith."  In column Z row 1 enter "fallnone."

Click on cell Y2 (Column Y row 2) to make it active.  From the "Insert" menu, select "Function."   In the "Paste Function" dialog box, click on "Logical in the "Function Category" field. Click on "If" in the "Function Name" field and then click "OK."  This box should appear as follows:

In the "Logical_test" field of the dialog box, enter D2=1" (D is the column in which the "falexper" is located, 2 is the second row number, and "1" specifies the condition to be satisfied).  In the "Value_if_true" field, enter X2 (to indicate that we wish the value in cell X2 to be placed in our new cell) and in the "value_if_false" field enter zero (Excel cannot handle missing values with the t test function--zeros must be inserted).

Click on cell Y2 to make it active.  On the "Edit" menu click on "Copy." Go to cell Y2 and hold down the left mouse button while moving down column Y until the end of the data is reached on row 53.  Press Return.

bulletRepeat the process for column z. Click on cell Z2 (column Z, row 2) to make it active. Click on "Insert" menu and then on "Function." In the "Paste Function" dialog box, click on "Logical" in the "Function Category" field. Click on "If" in the "Function Name" field and then click "OK." In the "Logical Test" field of the dialog box, enter "D2=2" (D is the column in which the "Falexper" variable is located, 2 is the second row number, and "=2" specifies the condition to be satisfied). In the "value_if_true" field enter Z2 (to indicate that we wish the value in cell Z2 to be placed in our new cell) and in the "value_if_false" field enter zero.

Click on cell Z2 to make it active. On the "Edit" menu click on "Copy." Go to cell Z2 and hold down the left mouse button while moving down column Y until the end of the data is reached. Press Return.

Excel cannot handle missing values in the t test function. Thus, students must sort the columns to create columns with all the data at the top for "FALEXPER" and "fallnone."

From the "Data" menu click "Sort."  Un the ""Sort" dialog box, click the "Sort by" field and enter "falexper." In the "followed by" field, enter "fallnone."  Click "descending" sorting by for each of the variables.  For "My list has . . .," click "header row."    Click "OK."

 

Image47.gif (5004 bytes)

This sort will create one column with seros above scores.  For this column, highlight all the zeros and click ont he "Edit" menu and click on "Delete."  On the "Delete" dialog box, click on "shift cells up" and click "OK."  Now you have all the data at the top of each column.

bulletTo compute t, locate a place to put the output. Go to row 56 column Y and click on the cell to make it active.

From the "Tools" menu click on "Data Analysis."  In the dialog box titled "Data Analysis" select t-Test: Two Sample Assuming Equal Variances" and click "OK."

In the dialog box "t-test: Two-Sample Assuming Equal Variances" go to the "Range" field, enter the starting cell and ending cell for the "fallnone" variable (in this case, $Z$2:$Z$30). Click on the "Output Range" button and enter the target cell in the field ($Y$56). Click "OK."

 

bulletSave your output and answer the following questions:

What is the observed t value?____________

Is there a significant difference between the groups?____________