Conducting t-tests in SPSS

Conducting t-tests in SPSS - One-sample t-test What it does...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: One-sample t-test What it does The one-sample t-test is used to determine whether or not the population mean of a dependent variable is equal to a particular value (called the "test value" in SPSS). Running the test in SPSS 1) 2) 3) 4) From the Analyze menu, select Compare Means, and then One-Sample T Test... . Move the dependent variable into the Test Variable box. Enter the desired number into the Test Value box. Click on OK. Interpreting the SPSS output The SPSS output first reports descriptive statistics: N, M, and SD. The entry labeled "Std Error Mean" (SEM in APA format) on the top line indicates how well the sample mean estimates the population mean; the smaller the SEM, the better the estimate. The sample mean probably differs from the test value, but is this difference statistically significant? To determine this, look at the second table in the SPSS output. This table reports the inferential statistics: t, df, and p, which is labeled "Sig. (2-tailed)." If p is greater than .05, the result is not statistically significant, and you should conclude that the population mean of the dependent variable is equal to the test value. However, if p is less than or equal to .05, the result is statistically significant and you should conclude that the population mean is not equal to the test value. If you have a significant result, you must determine its direction. For the one-sample t-test, this involves determining whether the population mean is above the test value or below the test value. (There is no need to do this if you do not have a significant result, because a non-significant result indicates that the population mean is neither above nor below the test value; it is equal to the test value.) If the sample mean is greater than the test value, assume that the population mean is also greater than the test value. However, if the sample mean is less than the test value, assume the population mean is below the test value as well. t-tests Page 1 of 6 One-sample t-test Reporting results in APA format 1) If you had a non-significant result, begin with this sentence: [Description of dependent variable] did not differ significantly from [test value]. If you had a significant result and M was greater than the test value, begin with this sentence: [Description of dependent variable] was significantly greater than [test value]. If you had a significant result and M was less than the test value, begin with this sentence: [Description of dependent variable] was significantly less than [test value]. Be sure you substitute appropriate wording for the description of the dependent variable, and enter the test value you used. An example would be IQ scores did not differ significantly from 100. 2) Correct the grammar of your sentence as needed. For example, if your sentence reads "Female scores was significantly higher than male scores," change this to "Female scores were significantly higher than male scores." 3) Report M and SD in parentheses immediately after you describe the dependent variable. The sentence might now read IQ scores (M = 102.25, SD = 8.38) did not differ significantly from 100. 4) Immediately before the period at the end of the sentence, insert a comma and this sequence: t([df]) = [value of t], p = [value of p]. Be sure you insert the appropriate values for df, t, and p. Here's an example: IQ scores (M = 102.25, SD = 8.38) did not differ significantly from 100, t(24) = 1.76, p = .09. 5) Round off M, SD, t, and p, so that you only report two digits to the right of the decimal point. Thus, if the third digit to the right of the decimal point is 5 or more, you should add 1 to the second digit to the right of the decimal point; otherwise, do not change the second digit. 6) If there is no digit to the left of the decimal point for M, SD, or t, put a zero before the decimal point. Do not put a zero before the decimal point for p. 7) If the value for t is negative, delete the minus sign in front of it. 8) If the end of your sentence reads p = .00, change it to p < .01. 9) Check to make sure you have left a space before and after each equals sign (=), and before and after each less than sign (<). t-tests Page 2 of 6 Paired-sample (dependent) t-test What it does The dependent t-test is used to determine whether or not two different dependent variables have the same population mean. Running the test in SPSS 1) 2) 3) 4) From the Analyze menu, select Compare Means, and then Paired-Samples T Test.... Select both variables to be compared by clicking on them one after the other. Click on the arrow to move both variables into the Paired Variables box. Click on OK. Interpreting the SPSS output In the top portion of the output, SPSS reports descriptive statistics (N, M, and SD) for each variable. Next, SPSS reports the correlation between the two variables. The final table lists the characteristics of the difference variable on the left side, and the inferential statistics on the right side. The latter consist of t, df, and p, which is labeled "Sig. (2-tailed)." The sample means of the two variables probably differ from each other, but is this difference statistically significant? To determine this, look at p, which is labeled "Sig. (2-tailed)." If p is greater than .05, the result is not statistically significant and you should conclude that the means of the two variables are equal to each other. However, if p is less than or equal to .05, the result is statistically significant and you should conclude that the means of the two variables are not equal to each other. If you have a significant result, you must determine its direction. For the dependent t-test, this involves determining which of the two variables has the larger population mean. (There is no need to do this if you do not have a significant result, because a non-significant result indicates that the population means of the two variables are equal to each other.) Assume that the variable with the larger sample mean also has the larger population mean. t-tests Page 3 of 6 Paired-sample (dependent) t-test Reporting results in APA format 1) If you had a non-significant result, begin with this sentence: [Describe one DV] did not differ significantly from [describe other DV]. If you had a significant result, begin with this sentence: [Describe DV with larger M] was significantly greater than [describe DV with smaller M]. Be sure you substitute appropriate wording for the descriptions of the two dependent variables. An example would be SAT math scores did not differ significantly from SAT verbal scores. 2) Correct the grammar of your sentence as needed. 3) Report the corresponding values of M and SD in parentheses immediately after you describe each dependent variable. The sentence might now read SAT math scores (M = 601.25, SD = 16.48) did not differ significantly from SAT verbal scores (M = 605.89, SD = 15.92). 4) Immediately before the period at the end of the sentence, insert a comma and this sequence: t([df]) = [value of t], p = [value of p]. Be sure you insert the appropriate values for df, t, and p. Here's an example: SAT math scores (M = 601.25, SD = 16.48) did not differ significantly from SAT verbal scores (M = 605.89, SD = 15.92), t(36) = 0.89, p = .38. 5) Round off M, SD, t, and p, so that you only report two digits to the right of the decimal point. Thus, if the third digit to the right of the decimal point is 5 or more, you should add 1 to the second digit to the right of the decimal point. 6) If there is no digit to the left of the decimal point for M, SD, or t, put a zero before the decimal point. Do not put a zero before the decimal point for p. 7) If the value for t is negative, delete the minus sign in front of it. 8) If the end of your sentence reads p = .00, change it to p < .01. 9) Check to make sure you have left a space before and after each equals sign (=), and before and after each less than sign (<). t-tests Page 4 of 6 Independent t-test What it does The independent t-test is used to determine whether or not two different groups have the same population mean for a single dependent variable. Running the test in SPSS 1) 2) 3) 4) 5) From the Analyze menu, select Compare Means, and then Independent-Samples T Test.... Move the dependent variable into the Test Variable(s) box. Move the variable that identifies the two different groups into the Grouping Variable box. Click on Define Groups... to make a new window appear. In the Group 1 box, if the grouping variable is numeric, type the number that identifies one of the two groups. If the grouping variable is a string variable, type the word that identifies one of the two groups. If you type a word, check your spelling and capitalization these must exactly match the entries used in the data grid. 6) In the Group 2 box, type the number or word that identifies the other group. 7) Click Continue. 8) Click OK. Interpreting the SPSS output In the top portion of the output, SPSS reports descriptive statistics (N, M, and SD) for each of the two groups. In the bottom portion of the output are the results of Levene's test (on the left side) and the inferential statistics (on the right side). For the inferential statistics t, df, and p, which is labeled "Sig. (2-tailed)" there are actually two sets of each of these numbers; one set is labeled "Equal variances assumed," while the other set of numbers is labeled "Equal variances not assumed." The sample means of the two groups probably differ from each other, but is this difference statistically significant? To determine this, look at p (labeled "2-Tail Sig" on the bottom line of the output). However, there are two p values. Which one should you use? Here's how to decide: 1) If the two groups are approximately equal in size, use the "Equal variances assumed" results (the second-to-last line in the output). 2) If the two groups differ markedly in size, examine the "Levene's Test" line. The p value for Levene's test is labeled "Sig." If p for the Levene's test is less than or equal to .05, use the "Equal variances not assumed" line (the bottom-most line of the output) to obtain the values of t, df, and p. Otherwise, use the "Equal variances assumed" (second-last) line. Once you've determined which line of the output to examine, check the value of p for the t-test, which is labeled "Sig. (2-tailed)." If p is greater than .05, you have a non-significant result and should conclude that the means of the two groups are equal to each other. However, if p is less than or equal to .05, you have a significant result and should conclude that the means of the two groups are not equal to each other. If you have a significant result, you must determine its direction. For the independent t-test, this involves determining which group has the larger population mean. (There is no need to do this if you do not have a significant result, because a non-significant result indicates that the population means of the two groups are equal to each other.) Assume that the group with the larger sample mean also has the larger population mean. t-tests Page 5 of 6 Independent t-test Reporting results in APA format 1) If you had a non-significant result, begin with this sentence: [Describe dependent variable] did not differ significantly between [describe one group] and [describe other group]. If you had a significant result, begin with this sentence: [Describe dependent variable] was significantly greater in [describe group with larger M] than in [describe group with smaller M]. Be sure you substitute appropriate wording for the descriptions of the two dependent variables. An example would be Reaction times were significantly greater in males than in females. 2) Correct the grammar of your sentence as needed. 3) Report the corresponding values of M and SD in parentheses immediately after you describe each dependent variable. The sentence might now read Reaction times were significantly greater in males (M = 351.25, SD = 6.48) than in females (M = 343.89, SD = 5.92). 4) Immediately before the period at the end of the sentence, insert a comma and this sequence: t([df]) = [value of t], p = [value of p]. Be sure you insert the appropriate values for df, t, and p. Here's an example: Reaction times were significantly greater in males (M = 351.25, SD = 6.48) than in females (M = 343.89, SD = 5.92), t(40) = 2.47, p = .02. 5) Round off M, SD, t, and p, so that you only report two digits to the right of the decimal point. Thus, if the third digit to the right of the decimal point is 5 or more, you should add 1 to the second digit to the right of the decimal point. 6) If there is no digit to the left of the decimal point for M, SD, or t, put a zero before the decimal point. Do not put a zero before the decimal point for p. 7) If the value for t is negative, delete the minus sign in front of it. 8) If the end of your sentence reads p = .00, change it to p < .01. 9) Check to make sure you have left a space before and after each equals sign (=), and before and after each less than sign (<). t-tests Page 6 of 6 ...
View Full Document

Ask a homework question - tutors are online