|
Chi-Square Test of Independence c2
PASW (SPSS) output for Regression Tests of Significance Normal distribution
PASW (SPSS) Instructions Exam questions More questions
PASW (SPSS) Instructions
Mean, Median Standard Deviation Variance etc
One Sample T test Test for Normality
Confidence Interval other than 95% and Lower and Upper quartiles
Paired Sample T-tests Independent Samples T-test
Regression line, Scatterplot and Residual Plots
One Sample T test
Analyze
Compare Means
One-Sample T Test...
Move the variable into test variable via the arrow.
Put the value of the Null Hypothesis in as the test value ie
Null Hypothesis: m = 24
Alternative Hypothesis m > 24
Please make sure you change it from zero as it will still do the test of significance
but it will be wrong unless the value in the Null Hypothesis is zero..
Back to the top
Test for Normality with one (or two independent samples)
Analyze
Descriptive Statistics
Explore
Put variable into Dependent list
(If 2 independent samples put Type into factor list, and Data into Dependent list)
Plots
Click Normality with plots
Continue
OK
Look at Shapiro-Wilk Sig if this is bigger than 0.05 then the data is normally distributed.
Back to the top
Confidence Interval other than 95% and lower and upper quartiles for one (and two independent samples)
Analyze
Descriptive Statistics
Explore
Put variable into Dependent list
(If 2 samples put Type into factor list and Data into Dependent list)
Statistics
Change Confidence Interval for mean to new value and tick percentiles
Continue
OK
Back to the top
Paired Sample T-tests
Transform
Compute Variable
Type the word ‘Difference’ in the Target Variable
Then Group 1 –Group 2
Continue
Now follow Test of Normality as before but with the variable ‘Difference’
Paired Sample T-tests
Analyze
Compare Means
Paired-Samples T Test
Move the variables into Paired Variable1 and Variable 2
( Note the way the difference was taken back in transform and do the same order here)
OK
Another method is to use one sample t-test with the difference
One Sample T test
Analyze
Compare Means
One-Sample T Test...
Move the variable, difference into test variable via the arrow.
Put the value of the Null Hypothesis in as the test value e.g.
Back to the top
Independent Samples T-test
In variable view
Name one column Data, this should be numeric
Name the other column Type or Group, this should be string,
Then in values you can put your letters and labels (you can put words such as with device) must press add after each one.
Please make sure that you put your data in the columns correctly.
One column should have all the data,
Second Column should be the groups
To check for Normality as before
Analyze
Compare Means
Independent-Samples T Test
Move the data variable to Test Variable
Move the group variable(Type) to Grouping Variable
Type(??) will appear,
Click Define groups
Put A (1) in Group 1
Put B (2) in Group 2
If you want the subtraction A - B, as SPSS will always take Group One - Group Two
Continue
Back to the top
Regression
Put the data into two columns
Analysis
Regression
Linear
Put the y-value in the dependent (this is the one to be predicted)
Put the x-value in the independent
Save
Residuals Unstandardised
Predicted ValueUnstandardised
For the Regression Line
Graph
Legacy Dialogs
Interactive
Scatterplot
Put the x-value horizontal Put the y-value vertical
Fit
Method
Regression
For the Residual plot
Graph
Legacy Dialogs
Interactive
Scatterplot
Put the x-value horizontal
Put the sresid vertical
Fit
Method
Mean
Back to the top
Mean, Median, Standard Deviation, Variance etc... (these are generated automatically in this command)
Analyze
Descriptive Statistics
Explore
Put variable into Dependent list
(If 2 independent samples put Type into factor list, and Data into Dependent list)
Plots
Tick or untick histogram, boxplot stem and leaf plot as required
Continue
OK
Here are the answers to the following questions
Tutorial Nine Question Two
Assumptions and Definitions
|
|
Kolmogorov- Smirnov
|
|
|
Shapiro-Wilk
|
|
|
|
|
Statistic
|
df
|
Sig.
|
Statistic
|
df
|
Sig.
|
|
Workload
|
.087
|
24
|
.200
|
..963
|
24
|
.521
|
|
|
Hypotheses
Null Hypothesis: m = 1
Alternative Hypothesis m ¹ 1
Testing the Null hypothesis that the true population workload measure is 1 against the alternative hypothesis that the population workload measure is now different to 1.
Test Data
Descriptives
|
|
|
|
Statistica
|
Standard error
|
|
Workload
|
Mean
|
|
1.2458
|
.1192
|
|
|
95% Confidence Interval for Mean
|
Lower Bound
|
.9993
|
|
|
|
|
Upper Bound
|
1.4924
|
|
|
|
5% Trimmed Mean
|
|
1.2509
|
|
|
|
Median
|
|
1.2000
|
|
|
|
Variance
|
|
.341
|
|
|
|
Std. Deviation
|
|
.5838
|
|
|
|
Minimum
|
|
.20
|
|
|
|
Maximum
|
|
2.20
|
|
|
|
Range
|
|
2.00
|
|
|
|
Interquartile Range
|
|
.9000
|
|
|
|
Skewness
|
|
-.176
|
.472
|
|
|
Kurtosis
|
|
-.872
|
.918
|
|
Test statistic
One-Sample Test
|
|
Test Value = 1
|
|
|
|
|
|
|
t
|
df
|
sig(2-tailed)
|
Mean difference
|
95%
Confidence
Interval of the
Difference
|
|
|
|
|
|
|
|
Lower
|
Upper
|
|
Workload
|
2.063
|
23
|
.051
|
.2458
|
-6.9428E-04
|
.4924
|
|
t 23 = 2.063 23 are the Degrees of Freedom (df)
P-value
We read this value directly from the table above as SPSS always does two sided tests.
Decision
So looking at our example we can see that the p-value is more than 0.05 so we will accept the Null Hypothesis.
Conclusion
The results are not significant indicating that there is insufficient evidence to suggest that workload is different to the base rate of 1 at 5% significance level.
|
