SPSS E xercise 12 Correlation and Box plot You are interested in learning about the relationship between

SPSS E xercise 12

Correlation and Box plot

You are interested in learning about the relationship between college grade point average, intelligence, SAT scores, introduction to psychology final exam scores and inorganic chemistry final exam scores. In order to describe this group, find all of the central tendency options, all of the dispersion options, and the quartiles. Use SPSS to create a correlation matrix for all of the variables (using Analyze, Correlate, Bivariate). Find the Pearson product moment correlation coefficient between all of the variables. Create a boxplot for the Psychology and the Chemistry exams.

GPA	IQ	SAT	Psychology	Chemistry
2.54	111	555	85	67
3.00	105	547	75	68
3.87	105	602	81	69
2.99	104	621	86	78
2.54	116	535	82	72
2.88	115	525	84	81
3.33	106	564	76	80
3.55	105	521	79	81
3.14	105	500	78	90
3.66	100	512	87	94
3.77	111	532	90	85
3.01	121	522	91	81
3.02	125	542	82	81
2.33	109	523	84	80
2.54	117	541	66	77
2.44	114	531	81	74
4.00	121	678	95	97
2.34	108	517	79	71
2.88	104	518	81	72
2.74	102	506	85	71

You now have graphs and statistics allowing you to describe your data. Before turning in the exercise, identify the following on the output:

Mean for the IQ variable

Standard deviation for the SAT variable

Median for the C hemistry variable

Mode for the Psychology variable

What is the correlation coefficient between GPA and Chemistry?

Is the correlation significant? Yes_____ No_____

What is the p value (level of significance)? __0.009_____

What is the correlation coefficient between GPA and SAT? ___0.456_____

Is the correlation significant? Yes_____ No_____

What is the p value (level of significance)? __0.043________

What is the correlation coefficient between SAT and Chemistry? ___0.183_____

Is the correlation significant? Yes_____ No_____

What is the p value (level of significance)? ____0.441______

Analyze the results of the correlation and boxplot and write a short explanation of your findings.

S PSS Exercise 13

Reliability Analysis

A measure is reliable if it produces consistent scores across administrations. Many of you will be in situations where you will have data from tests or questionnaires and you will want to determine the reliability of the instrument you are using. One way of doing this is by conducting a type of reliability analysis called "internal consistency". Internal consistency is an estimate of reliability based on the administration of a measure containing multiple parts, with the measure being administered to individuals on a single occasion.

High correlation coefficients mean lots of agreement thus high reliability. Low correlations mean lots of disagreement and thus low reliability. Fortunately, the correlation coefficient gives us the degree of agreement between the test items. In the case of the social sciences, an r of .80 or higher is needed to conclude that a measure is reliable.

In all cases of internal consistency estimates the instrument’s scores are summed. You may also be faced a situation where you might have to transform (convert to z scores or ‘reverse’ score items) items before they can be summed. For the purpose of this exercise, we will be working with data that do not require any transformations of the items. The data for this analysis is a 10-item questionnaire measuring test anxiety and we want to answer the question "How reliable is this 10-item measure of test anxiety?"

Open SPSS and name the variable and then enter the data for each question in the questionnaire as shown in the table below. Click on ANALYZE and then on SCALE. Select RELIABILITY ANALYSIS. When you see the dialog box, make sure Alpha is selected in the Model box and not one of the other reliability methods. Move all of the question items from the left box to the right box in the usual manner. Click on STATISTICS and check on item and scale in the D escriptives for area and then click on c orrelations in the i nter -I tem area . Then click on CONTINUE and then OK. Your reliability analysis should appear in your output window.

Participant	q1	q2	q3	q4	q5	q6	q7	q8	q9	q10
1	2	3	1	2	1	2	3	2	3	3
2	3	3	2	3	4	3	2	3	5	5
3	2	5	2	1	2	2	2	2	1	3
4	3	5	3	5	5	5	5	4	5	5
5	2	3	2	3	3	4	4	3	3	3
6	1	3	1	2	3	3	5	3	2	3
7	2	2	2	1	1	1	2	2	1	1
8	2	5	3	2	5	3	3	3	1	3
9	3	4	3	2	2	2	2	4	2	3
10	2	3	3	2	2	5	5	4	4	4

Print your results (change the printer output to landscape to keep the correlation matrix on one page) and identify the following on the output:

Coefficient alpha _________
Correlation between questions 4 and 5 __________
Correlation between questions 6 and 7 __________
Mean for question 9 ____
Standard deviation for question 8 ____ 0.8165 ______

What can be said about the overall scale reliability for our survey on test anxiety? Write a brief results section to report your analyses.

SPSS Exercise 14a & b

Regression

In this exercise, you will be using the academic data that was given with the correlation exercise. You will find the basic descriptive information about the chemistry and psychology tests. Next you will find the equation that best represents the linear relationship between variables X (chemistry) and Y (psychology) and determine the strength of the relationship between chemistry and psychology test scores.

Open the academic data that you used in the correlation exercise. Select "Analyze" and then "Regression" and then "Linear". In this case, in the dialog box, select psychology for the dependent variable and chemistry for the independent variable. Next click "Statistics" in the dialog box and click in the box for "descriptives". Click on "Continue" and then OK.

Next, create a scatterplot. On the menu bar select "Graphs" and then click on "scatter" Click on "Simple" and then the "Define" button. Click on Psychology for the Y-axis and chemistry for the X-axis. Click on titles and in rectangle number one type-"scatterplot to predict psychology from chemistry" Click "continue" and then ok.

Finally, fit a regression line to the scatterplot by ‘double-clicking’ on the scatterplot, then place the mouse pointer on one of the circles (any one will do) and click on the circle (this action will highlight all of the circles in the scatterplot). Scroll your mouse pointer over the graphs in the menu toolbar until you see "Add Fit Line at Total". Click on the graph and you will see another window open (Properties). Close the properties window and then close the chart editor and you will have a regression line fitted to your data.

Print out your data. Find the information needed to solve the problem (look at the output and find the section that contains the ‘coefficients’. Once you locate this section locate the constant under the ‘Unstandardized Coefficients’ column. This number represents the a in the regression equation. Then locate the number associated with the ‘Chemistry’ variable in the same column. This number represents the b in the regression equation.

SPSS Exercise 14b

Sally Sage has a score of 81 in chemistry. Using the information from the output for the problem above, write out the regression equation and predict Sally’s psychology test score (this last part is a hand calculation).

SPSS Exercise 15

Chi-Square Two-Way Contingency Table Analysis

Chi-square ( ${{\chi }^{2}}$ ) is a technique for nominal data to compare observed data with data that is expected. The SPSS program will create a table of your data and will determine if the information in the table is a chance pattern or something else.

In this case we are looking at the restaurant preferences of males and females. Their choices are Chinese, Italian or French restaurants. The data are entered under the variable names of Gender and Food (Figure 1). For the Gender variable, in the labels column enter 1 = male 2 = female and in the Food variable, in the labels column enter 1 = Chinese; 2 = Italian; 3 = French. (Remember that you have as an alternative to change the variable from numeric to string and then you can type in the words male, female and Chinese, Italian, French) Enter the given data (Figure 2).

In order to run the Chi-square for this type of problem, select "Analyze" from the menu bar and the "Descriptive Statistics" and then "Crosstabs". In the cross tabs dialog box, select the gender variable for the row variable and the food variable for the column variable. Click on the "Cells" box and then in the "Counts" section place a check in the box by the word "expected" (Observed should already be checked) so that your table will show the expected cell values as well as the observed data that you entered, then in the "Percentages" section, click on "Row", "Column", and "Total" so that your table will show the percentages for each category. Click the "Continue" box. Now, click on the "Statistics" box. Place check marks in the box next to "Chi square" and "Phi coefficient and Cramer's V". Click continue and then OK.

id	gender	food	id	gender	food
1	1	1	22	2	3
2	1	2	23	2	3
3	1	2	24	2	1
4	1	2	25	2	2
5	1	3	26	2	3
6	1	3	27	2	3
7	1	2	28	2	3
8	1	1	29	2	3
9	1	3	30	2	3
10	1	1	31	2	1
11	1	2	32	2	1
12	1	2	33	2	1
13	1	2	34	2	2
14	1	2	35	2	2
15	1	2	36	2	2
16	1	1	37	2	3
17	1	2	38	2	3
18	1	2	39	2	3
19	1	3	40	2	3
20	2	3	41	2	3
21	2	1	42	2	3

Identify the appropriate Chi Square analysis for this question (circle or highlight the Correct test) Test of Independence Goodness of Fit
State the null and alternative hypotheses
H _o :
H ₁ :
Conduct a crosstabs analysis to examine whether there are differences between restaurant preferences and gender (male/female). From the output, identify the following:
1. Percent of females who’s restaurant preference was Chinese
2. Percent of males who’s restaurant preference was French
3. Total percent of males and females who’s restaurant preference was Italian
4. ² value
5. Strength of relationship between gender and restaurant preference
  (phi’s of .10, .30, and .50 are considered to be small, medium and large, respectively)
Create a clustered bar graph to show the differences between gender and restaurant preference
Are gender and food preference related? Discuss your findings