File you will need for this assignment:
Data from the survey
Copy of the survey
Each of you will be assigned a research question to answer from data that a class collected from a survey. You will need to run a t-test to answer the research question. The t-test is appropriate because the independent variable has just two levels (i.e., male and female). I have created a spreadsheet to calculate the t-test for you and a PowerPoint presentation that describes how to use the spreadsheet. Enter the independent variable in the IV column and the dependent variable in the DV column. The spreadsheet will do the rest. The spreadsheet is set for one group to have been labeled 1 and the other group to have been labeled 2.
Type your assignment similar to the sample below.
For this assignment we have set the alpha level (p) at .05. You need to decide whether to read the equal or unequal variance t-test on the spreadsheet. If the F-max test is less than or equal to .05 you read the unequal. If the F-max is greater than .05, you read the equal variance t-test.
If the p (two tailed significance) for your t-test is less than or equal to .05, you will reject your null hypothesis and state that there is a difference between the means of the two groups. If your p > .05, you will fail to reject your null hypothesis and state that there is no difference between the means of the two groups.
You also need to state the effect size. EFFECT SIZE is used to calculate practical difference. Effect size is the difference of the two means divided by the standard deviation of the control group (or the average standard deviation of both groups if you do not have a control group). Effect size becomes important if you have statistical significance. An effect size of .2 is considered small, .5 is considered medium, and .8 is considered large. The spreadsheet program will calculate the EFFECT SIZE for you (you just need to select the correct one).
The number in () following the t is the degree of freedom (number of subjects minus the number of groups. The number following the = is the t-value. In the example below, there are 32 degree of freedom (number of subjects minus the number of groups) and the t-value is 4.58. n is the number of subjects. In the example below there are 17 females and 17 males. p is the significance level (a.k.a. probability -- the likelihood of finding this difference between the means of the two groups by chance). Although it is not written in stone, p < .05 is a commonly used. Don't forget to underline M, SD, p, t, and n. We also underline the title of the table and put a period after the t statement that is under the table. We only say there is a difference in the means of the two groups if the t-value is statistically significant (p < .05). If p is not less than .05 we still report the means for the two groups, but we say they are similar. We will only report the effect size (d) if there is a significant difference between the mean. We will not report it if there is not a significant difference because that would mean that the effect size probably occurred by chance. I have listed the scoring system throughout the sample assignment in red.
Sample Assignment (yours will be different , but have these elements)....
Study Design
Independent variable: Student Gender
Dependent variable: Mathematics Achievement (4 points)
Type of t-Test and Why
Equal variance independent t-test because the subjects in the two groups are different and while the number of subjects in each group is different, the variance of the two groups is similar. (3 points)
Question
Is there a significant difference between boys
and girls with respect to mathematics achievement?
--or--
Does mathematics achievement differ between
boys and girls? (This will be provided for you.)
Null Hypothesis
Mathematics achievement does not differ between
boys and girls. (1 point)
Alternative Hypothesis
Mathematics achievement differs between boys
and girls. (1 point)
Answer
There was a significant difference in the
mathematics achievement of boys and girls, t(32)=4.58, p=.04. This
difference represented a large effect size, d=3.73 (Cohen, 1988). Girls (M=12.3,
SD=.56) scored lower than boys (M=14.5, SD=.62). (1 point for correct answer, 1 point for underline t,
1 point for correct degree of freedom, 1 point for correct t-value, 1 point for underline
p, 1 point for correct p value, 1 point for underline d, 1 point for correct
effect size, 2 points for underline M (twice), 2 points for correct mean values, 2
points for underline SD (twice), and 2 points for correct standard deviation
values)
Table 1
Gender Differences in Mathematics Achievement
------------------------------------------------------------------------------------------------
Gender M SD n
-------------------------------------------------------------------------------------------------
Females 12.3 .56 17
Males 14.5 .62 17
-------------------------------------------------------------------------------------------------
t(32)=4.58, p=.04, d = 3.73.
(1 point for Table 1, 1 point for table title, 1 point for underline table title, 1 point for line (rule) above category head, 1 point for independent variable name in category head, 1 point for underline M, 1 point for underline SD, one point for underline n, 1 point for line (rule) under category head, 2 points for two levels of the IV, 2 points for correct means, 2 points for correct standard deviations, 2 points for correct number of subjects in each group, 1 point for bottom line (rule), 1 point for underline t, 1 point for correct degree of freedom, 1 point for correct t-value, 1 point for underline p, 1 point for correct p value, 1 point for including value of d, 1 point for underlining d, and 1 point for period at end of d statement.)
Del Siegle
del.siegle@uconn.edu