SPSS 21 Syntax Cheat Sheet

=====================================

This article provides an overview of the key syntax commands and functions available in SPSS version 21 for performing various statistical analyses.

### Data Preparation

#### Creating Variables

To create a new variable called `newvar` based on existing variables `x1`, `x2`, etc. use:

```spss

compute newvar = x1 + (abs(x2) > 0)*x2.

vif newvar.

```

Alternatively, you can also use the following syntax to create a new variable that takes only positive values from another variable:

```spbs

genlog newvar = abs(varname).

exe.

```

### Descriptive Statistics

To calculate basic summary statistics (mean, standard deviation) for continuous variables `x1`, `x2` etc., use the following syntax in SPSS 21:

```r

descript x1-x5 / mean std.

```

For categorical variables or binary outcomes you can also specify additional options such as frequencies and missing values.

### Frequency Analysis

To create frequency tables for nominal/ordinal/categorical variables `varname`, you can use the following syntax in SPSS 21:

```spss

freq varname / table cellcount .

```

This will display a simple count of unique levels along with their corresponding frequencies. You may also want to specify additional options such as collapsing categories or displaying marginal probabilities.

### T-Test and ANOVA

To perform an independent t-test on continuous variables `x1` vs `x2`, you can use the following syntax in SPSS 21:

```spss

ttest x1 = x2.

```

For within-subjects repeated measures designs, you may want to use a related samples t-test instead.

To perform an ANOVA (Analysis of Variance) on continuous outcome variable `y` with one or more independent variables (`x1`, `x2`) and any interaction terms, you can use the following syntax in SPSS 21:

```spss

anovapwer y by x1 * 'x2'.

```

### Linear Regression

To perform a linear regression analysis of continuous outcome variable `y` on one or more independent variables (`x1`, `x2`) and any interaction terms, you can use the following syntax in SPSS 21:

```spss

regresison y = x1 + 'x2' * (x1*x2).

```

### Correlation Analysis

To calculate Pearson's correlation coefficient between two continuous variables `var1` and `var2`, you can use the following syntax in SPSPS 21:

```r

corr matrix=matrixname.

```

This will produce a square correlation matrix with all possible pairs of variables on the diagonal.

### Non-Parametric Tests

For non-normal or ordinal data, it is often more appropriate to use non-parametric alternatives such as Kruskal-Wallis H test for k independent groups or Wilcoxon's rank-sum (Mann-Whitney U) tests. These can be performed using similar syntax as their parametric counterparts but with `nonparam:kruskall` and `nonparam:wilcoxon` options respectively.

### Missing Data Handling

Missing values are typically denoted by a period (`.`). You may want to explore missingness patterns before proceeding with any further analysis. There are several ways to handle missing data in SPSS, including:

* Listwise deletion: This involves deleting all cases where there is even one missing value for the variable(s) of interest.

* Pair-wise or casewise deletion: These involve deleting entire rows (cases) if either all values or any single pair/observation respectively are missing.

### Additional Resources

For more detailed information on SPSS syntax and functions, please refer to the official IBM documentation:

https://www.ibm.com/support/knowledgecenter/en/3C8vycnspss-24_0pro_fcs_index.htm

Spss 21 syntax cheat sheet