SPSS SKILLS Flashcards
z-test
even though spss is not able to perform a z test, it can be achieved by performing a one sample - t test
1) analyze
2) compare means
3) one sample t test
4) move variables to test variables
one sample t- test
1) analyze
2) compare means
3) one sample t-test
4) select your test variable
5) if you want to test for you hypothesis that the “average value is ekis” put that value on test value box
6) paste
7) mean column indicates the mean of the sample
8) mean difference column indicates the difference between mean of sample and the expected mean from your hypotheses (the one being tested)
9) confidence interval of the difference column - indicates the confidence between the sample average and the hypothesized average for the population
to find the confidence level you need to rest both lower and upper columns from the hypothesized average
e.g. lower = -1.4
upper = -.68
test value (hypothesized) = 5.5
5.5 - 1.4 = 4.1
5.5 - 0.7 = 4.8
hence, confidence interval lies between 4.1 and 4.8
paired sample t-test
1) analyze
2) compare means
3) paired sales t-test
4) insert pre and post measurements in box, post as variable 1 and pre as variable 2
5) paste and run
independent samples t-test
1) analyze
2) compare means
3) independent samples t-test
4) select your variables for which you want to compare means
5) identify the groups you want to compare (coding for variables 0,1)
6) paste and run
f-ratio
1) analyze
2) compare means
3) one way ANOVA
4) move variables to factor and dependent list boxes
5) options
6) descriptive
7) continue
one-way ANOVA
1) analyze
2) compare means
3) one way ANOVA
4) place variables
5) post hoc - bonferroni
6) options - descriptives - homogeneity of variance - means plot
7) paste and run
interaction in ANOVA in SPSS
1) analyze
2) general linear model
3) univariate command
4) numerical variable selected as dependent variable
5) categorical variables are selected as fixed factors
6) plots
7) put factors on horizontal axis vs separate lines
8) post hoc - bonferroni
9) options - descriptives - homogeneity
correlation coefficient
1) analyze
2) correlate
3) bivariate
4) move variables to box
5) click Pearson and two tailed
6) paste and run
regression line on SPSS
1) analyze
2) regression
3) linear
4) move your dependent variable (variable being predicted) to the dependent variable box
5) move variable to independent variable box
6) graphs
7) legacy dialogs
8) scatter/dots
9) simple scatter and define
10) click your variable and move it to the variable label to the Y axis box
11) click your other variable and move it to the variable label to the X axis box
12) double click chart to select for editing
13) add fit line at total button
correlation analysis
1) graphs
2) legacy dialogues
3) scatter dots
4) simple scatter
5) select your variables on axis
6) paste and run
7) double click
8) add fit line at total
9) linear
10) dont attach label to line
11) correlate
12) bivariate
13) select variables you want to correlate
14) check Pearson
15) check spearman, paste, run
simple regression
1) analyze
2) regression
3) linear
4) select dependent variable and predictor/independent
5) statistics - confidence intervals
6) paste, run
R square- convert to percentage and that is the percentage the regression model explains of the dependent variable
multiple regression
1) analyze
2) regression
3) linear
4) add your predictors as independent variables
5) add your outcome variable as dependent variable
6) statistics
7) confidence interval
8) paste and run
- model summary shows R and R square
- R square represents the % of the variance in your outcome that can be explained by the model with your predictors
- ANOVA table - shows regression and residual
- constant represent where the lines cross the y axis
regression analysis with a dummy variable
1) Transform > Recode into Different Variables.
2) Select the categorical variable you want to recode and move it into the box.
3) Click Old and New Values.
4) Enter the value for one category (e.g., “Male”) in the Old Value field and set its new value as 1.
5) Enter the value for the other category (e.g., “Female”) in the Old Value field and set its new value as 0.
6)Click Continue, then name the new variable in the output box.
Click OK to create the new dummy variable.
run regression analysis with dummy variable (independent box)
Chi square analysis with a contingency table on SPSS:
Analyze
Descriptive statistics
Cross tabs
Independent variable - columns
Dependent variable - rows
Cells
Counts - observed and expected
Percentages - columns
Statistics - chi square
Lambda
Paste and run
We are interested in the goodman and kruskal tau
Value of dependent - convert to % and that % is what the dv improved after taking into account the iv
chi-square requirements
Nominal + symmetrical/non directional: cramer’s V*, Phi**
Nominal + asymmetric/directional: Goodman& Kruskal’s Tau
Ordinal + symmetrical/non directional: gamma
Ordinal + asymmetric/directional: Somer’s D
