SPSS Complex Samples - data analysis

Everything You Need for Data Analysis

As a researcher, you want to be confident about your results. Performing data analysis in SPSS Complex Samples helps you to achieve more statistically valid inferences for populations measured in your complex sample data. SPSS Complex Samples provides you with better results because, unlike most conventional statistical software, it incorporates the sample design into survey analysis. And, it easily plugs into SPSS Base so you can seamlessly work in the SPSS environment.

SPSS Complex Samples provides you with four procedures to analyze data from sample survey data.

Complex Samples Descriptives (CSDESCRIPTIVES)—Estimates means, sums and ratios, and computes standard errors, design effects, confidence intervals hypothesis tests for samples drawn by complex methods. The procedure estimates variances by taking into account the sample design used to select the sample, including equal probability and probability proportionate to size (PPS) methods, and without replacement (WOR) sampling procedures. Optionally, CSDESCRIPTIVES performs analyses for subpopulations.

You can also use CSDESCRIPTIVES to specify how to handle missing data:

Base each statistic on all valid data for the analysis
variable(s) used in computing the statistic. Compute ratios using all cases with valid data for both of the specified variables. You may base statistics for different variables on different sample sizes.
Base only cases with valid data for all analysis variables when computing statistics. Always base statistics for different variables on the same sample size.
Exclude user-missing values among the strata, cluster and subpopulation variables
Include user-missing values among the strata, cluster and subpopulation variables. Treat user-missing values for these variables as valid data.

Complex Sample Tabulate (CSTABULATE)—Displays one-way frequency tables or two-way crosstabulations and associated standard errors, design effects, confidence intervals and hypothesis tests, for samples drawn by complex sampling methods. The procedure estimates variances by taking into account the sample design used to select the sample, including equal probability and PPS methods, and with replacement (WR) and WOR sampling procedures. Optionally, CSTABULATE creates tables for subpopulations.

Use the following statistics within the table:

Population size
Standard error
Row percentages
Column percentages
Table percentages
Coefficient of variation
Design effects
Square root of the design effects
Confidence interval
Unweighted counts
Cumulative population size estimates
Cumulative percentages
Expected population size estimates
Pearson residuals
Adjusted Pearson residuals

Use the following statistics and tests for the entire table:

Test of homogeneous proportions
Test of independence
Odds ratio
Relative risk
Risk difference

Like CSDESCRIPTIVES, you can also use CSTABULATE to specify how to handle missing data. You can:

Base each table on all valid data for the tabulation variable(s) used in creating the table. You may base tables for different variables on different sample sizes.
Use only cases with valid data for all tabulation variables in creating the tables. Always base tables for different variables on the same sample size.
Exclude user-missing values among the strata, cluster and subpopulation variables
Include user-missing values among the strata, cluster and subpopulation variables. Treat user-missing values for these variables as valid data.

Complex Samples General Linear Models (CSGLM)—Enables you to build linear regression, analysis of variance (ANOVA), and analysis of covariance (ANCOVA) models for samples drawn by complex sampling methods. The procedure estimates variances by taking into account the sample design used to select the sample, including equal probability and PPS methods, and WR and WOR sampling procedures. Optionally, CSGLM performs analyses for subpopulations.

You can use the following statistics with CSGLM:

Model parameters: Coefficient estimates, standard error for each coefficient estimate, t test for each coefficient estimate, confidence interval for each coefficient estimate, and square root of the design effect for each coefficient estimate
Population means of dependent variable and covariates
Model fit
Sample design information

Hypothesis tests include:

Test statistics: Wald F test, adjusted Wald F test, Wald Chi-square test, and adjusted Wald Chi-square test
Adjustment for multiple comparisons: Least significant difference, Bonferroni, sequential Bonferroni, Sidak, and sequential Sidak
Sampling degrees of freedom: Based on sample design or fixed by user
Estimated means: Request eliminated marginal means for factors and interactions in the model

Handle missing data using listwise deletion of missing values.

Complex Samples Logistic Regression (CSLOGISTIC)—Performs binary logistic regression analysis, as well as multiple logistic regression (MLR) analysis, for samples drawn by complex sampling methods. The procedure estimates variances by taking into account the sample design used to select the sample, including equal probability and PPS methods, and WR and WOR sampling procedures. Optionally, CSLOGISTIC performs analyses for subpopulations.

You can use the following statistics with CSLOGISTIC:

Model parameters: Coefficient estimates, exponential estimates, standard error for each coefficient estimate, t test for each coefficient estimate, confidence interval for each coefficient estimate, design effect for each coefficient estimate, square root of the design effect for each coefficient estimate, covariances of parameter estimates, and correlations of the parameter estimates
Model fit: Pseudo R-squared and classification table
Summary statistics for model variables
Sample design information

Hypothesis tests include:

Test statistics: Wald F test, adjusted Wald F test, Wald Chi-square test, and adjusted Wald Chi-square test
Adjustment for multiple comparisons: Least significant difference, Bonferroni, sequential Bonferroni, Sidak, and sequential Sidak
Sampling degrees of freedom: Based on sample design or fixed by user

Handle missing data using listwise deletion of missing values.

Download the SPSS Complex Samples spec sheet (PDF) for more information on the features in SPSS Complex Samples.
Find more products for data analysis and other stages in the analytical process.

Buy Now

SPSS Complex Samples™

Everything You Need for Data Analysis