The document discusses the dependent samples t-test, a statistical method used to compare two sets of scores from the same sample. It outlines the steps for conducting the test, including calculating difference scores and applying the t-test formula, as well as providing examples and scenarios where this test is applicable. Additionally, it includes guidance on using SPSS for both one-sample and dependent samples t-tests.
The document discusses the dependent samples t-test, a statistical method used to compare two sets of scores from the same sample. It outlines the steps for conducting the test, including calculating difference scores and applying the t-test formula, as well as providing examples and scenarios where this test is applicable. Additionally, it includes guidance on using SPSS for both one-sample and dependent samples t-tests.
The document discusses the dependent samples t-test, a statistical method used to compare two sets of scores from the same sample. It outlines the steps for conducting the test, including calculating difference scores and applying the t-test formula, as well as providing examples and scenarios where this test is applicable. Additionally, it includes guidance on using SPSS for both one-sample and dependent samples t-tests.
The document discusses the dependent samples t-test, a statistical method used to compare two sets of scores from the same sample. It outlines the steps for conducting the test, including calculating difference scores and applying the t-test formula, as well as providing examples and scenarios where this test is applicable. Additionally, it includes guidance on using SPSS for both one-sample and dependent samples t-tests.
Download as PPTX, PDF, TXT or read online from Scribd
Download as pptx, pdf, or txt
You are on page 1/ 39
Dependent Samples T Test
25/11/2024 How many tests are there now??? • Z test we know everything about a population of interest o μ, o We are comparing a single sample’s mean against a population mean o M, n • One-sample t-test we only know the mean of a population o only μ o We are still comparing a single sample to that population, with additional estimations o M, n, df, s* Previously on One-sample T-Test
• Suppose your college newspaper reports a survey showing that students at your college study an average of 17 hours per week. However, you think that the students in your dormitory are much more hardworking than that. You randomly ask 5 students from your dormitory how much they study each week. They reported as below: 16, 18, 20, 20, 21 Previously on One-sample T-Test
• Suppose your college newspaper reports a survey showing that students at your college study an average of 17 hours per week. However, you think that the students in your dormitory are much more hardworking than that. You randomly ask 5 students from your dormitory how much they study each week. They reported as below: 16, 18, 20, 20, 21 o μ = 17, M =19, s = 2, df = 4 o t(4) = +2.25, p < .05 Which test should we use? Which test should we use?
• An average baby starts walking around 14 months, with a standard deviation
of 3 months. Researchers hypothesize that sending infants to swim lessons leads to an earlier walking onset. They studied 64 infants who have been attending swimming lessons in a local swimming school and found that the average infant in that school started walking around 12 months. Which test should we use?
• A company wants to test that whether their newly developed batteries last more than the average phone battery in the market, which is about 40 hours. They randomly selected 25 phones with the new battery and run them until they died, which took 45 hours on average. Which test should we use? • Four METU students are worried that they are living sedentary lives. The average daily steps they took last month are below: 6800, 6300, 6000, 4900. • They think this is not enough and promise each other to increase their average daily steps. Two months later, the average daily steps they took are as below: 6800, 7000, 5800, 5400 • Did they keep their promise? • In most research, we do not know the population’s mean either… • A more common research situation is studying the same sample across time, or for multiple variables at once • Repeated measures design: research strategy in which each person is tested more than once for the same variable(s) o e.g., before and after some psychological or social intervention • Within-subjects design: research strategy in which multiple variables were measured for each person o Virtually the same thing Dependent-samples T-test
• The hypothesis-testing procedure in which two sets of scores from the same sample is being compared Steps-- Steps-- Before After 1 6800 6800
2 6300 7000
3 6000 5800
4 4900 5400 Dependent-samples T-test
• Two scores from each person
o t test for dependent means o related samples t test o repeated measures t test o Paired-samples t test… • Determines the significance using difference or change scores • Everything is exactly the same way as a one-sample t test, except that o (a) you use difference scores o (b) the population mean will always be 0 Dependent-samples T-test
• Reactions to two different stimuli, performance on two different tasks • Developmental or longitudinal studies • Also used when participants are matched or paired based on theoretical reasons o Twins o Spouses o Brain-damage studies Dependent-samples T-test
• Clinical psychologists working in the APA tested a new type of therapy
for anxiety. They randomly selected 5 incoming patients to receive the new therapy. The patients filled out a standard anxiety scale before and after receiving the therapy. Anxiety-Before Anxiety- After
*) Determine which test is appropriate: if the population standard deviation is not known but the population mean is, go for one-sample t-test
1) Determine the population mean
2) Calculate the sample mean, if not provided
3) Estimate* the standard deviation, if not provided
4) Apply the t-test formula
5) Determine if the hypothesis is directional (cut-off p < .05) or non-directional (cut-off p < .025)
6) Find the cut-off t-value using the degrees of freedom (n-1) & cut-off p value
7) If the observed t values exceeds the cut-off t-value, reject the null hypothesis research hypothesis is supported Dependent sample t-test - Step by step
*) Determine which test is appropriate: if there is no mention of population, but the research is conducted as within-subjects design; go for a dependent samples t-test
1) Determine difference scores
2) Calculate the mean difference,
3) Estimate* the standard deviation of difference scores
4) Apply the t-test formula where population mean is 0
5) Determine if the hypothesis is directional (cut-off p < .05) or non-directional (cut-off p < .025)
6) Find the cut-off t-value using the degrees of freedom (n-1) & cut-off p value
7) If the observed t values exceeds the cut-off t-value, reject the null hypothesis research hypothesis is supported • SPSS guide to one-sample t-test --Analyze Compare Means One Sample T test --Move the target variable into the “Test Variable(s)” box --Replace the “0” in the box that says “Test value” with the population mean you want to test your sample against --Press OK • SPSS guide to dependent sample t-test --Analyze Compare Means Paired Samples T test --Move the target variable into the “Test Variable(s)” box --Place your variables (e.g., before x, after x) under the boxes labeled “Variable 1” and “Variable 2” --Press OK c c
