Values
Values
IBM
Note
Before using this information and the product it supports, read the information in “Notices” on page 19.
Product Information
This edition applies to version 26, release 0, modification 0 of IBM SPSS Statistics and to all subsequent releases and
modifications until otherwise indicated in new editions.
Contents
Missing values . . . . . . . . . . . . 1 MULTIPLE IMPUTATION Command Additional
Introduction to Missing Values . . . . . . . . 1 Features . . . . . . . . . . . . . . 12
Missing Value Analysis . . . . . . . . . . . 1 Working with multiple imputation data . . . . 12
Displaying Patterns of Missing Values . . . . . 3 Analyzing Multiple Imputation Data . . . . . 13
Displaying Descriptive Statistics for Missing Multiple Imputation Options . . . . . . . 16
Values . . . . . . . . . . . . . . . 4
Estimating Statistics and Imputing Missing Values 5 Notices . . . . . . . . . . . . . . 19
MVA Command Additional Features . . . . . 7 Trademarks . . . . . . . . . . . . . . 21
Multiple Imputation . . . . . . . . . . . . 7
Analyze Patterns . . . . . . . . . . . . 8 Index . . . . . . . . . . . . . . . 23
Impute Missing Data Values . . . . . . . . 9
iii
iv IBM SPSS Missing Values 26
Missing values
The following missing values features are included in SPSS® Statistics Premium Edition or the Missing
Values option.
Missing Values provides two sets of procedures for handling missing values:
v The Multiple Imputation procedures provide analysis of patterns of missing data, geared toward
eventual multiple imputation of missing values. That is, multiple versions of the dataset are produced,
each containing its own set of imputed values. When statistical analyses are performed, the parameter
estimates for all of the imputed datasets are pooled, providing estimates that are generally more
accurate than they would be with only one imputation.
v Missing Value Analysis provides a slightly different set of descriptive tools for analyzing missing data
(most particularly Little's MCAR test), and includes a variety of single imputation methods. Note that
multiple imputation is generally considered to be superior to single imputation.
You can get started with analysis of missing values by following these basic steps:
1. Examine missingness. Use Missing Value Analysis and Analyze Patterns to explore patterns of
missing values in your data and determine whether multiple imputation is necessary.
2. Impute missing values. Use Impute Missing Data Values to multiply impute missing values.
3. Analyze "complete" data. Use any procedure that supports multiple imputation data. See “Analyzing
Multiple Imputation Data” on page 13 for information on analyzing multiple imputation datasets and
a list of procedures which support these data.
Missing value analysis helps address several concerns caused by incomplete data. If cases with missing
values are systematically different from cases without missing values, the results can be misleading. Also,
missing data may reduce the precision of calculated statistics because there is less information than
Example. In evaluating a treatment for leukemia, several variables are measured. However, not all
measurements are available for every patient. The patterns of missing data are displayed, tabulated, and
found to be random. An EM analysis is used to estimate the means, correlations, and covariances. It is
also used to determine that the data are missing completely at random. Missing values are then replaced
by imputed values and saved into a new data file for further analysis.
Statistics. Univariate statistics, including number of nonmissing values, mean, standard deviation,
number of missing values, and number of extreme values. Estimated means, covariance matrix, and
correlation matrix, using listwise, pairwise, EM, or regression methods. Little's MCAR test with EM
results. Summary of means by various methods. For groups defined by missing versus nonmissing
values: t tests. For all variables: missing value patterns displayed cases-by-variables.
Data considerations
Data. Data can be categorical or quantitative (scale or continuous). However, you can estimate statistics
and impute missing data only for the quantitative variables. For each variable, missing values that are not
coded as system-missing must be defined as user-missing. For example, if a questionnaire item has the
response Don't know coded as 5 and you want to treat it as missing, the item should have 5 coded as a
user-missing value.
Frequency weights. Frequency (replication) weights are honored by this procedure. Cases with negative
or zero replication weight value are ignored. Noninteger weights are truncated.
Assumptions. Listwise, pairwise, and regression estimation depend on the assumption that the pattern of
missing values does not depend on the data values. (This condition is known as missing completely at
random, or MCAR.) Therefore, all methods (including the EM method) for estimation give consistent and
unbiased estimates of the correlations and covariances when the data are MCAR. Violation of the MCAR
assumption can lead to biased estimates produced by the listwise, pairwise, and regression methods. If
the data are not MCAR, you need to use EM estimation.
EM estimation depends on the assumption that the pattern of missing data is related to the observed data
only. (This condition is called missing at random, or MAR.) This assumption allows estimates to be
adjusted using available information. For example, in a study of education and income, the subjects with
low education may have more missing income values. In this case, the data are MAR, not MCAR. In
other words, for MAR, the probability that income is recorded depends on the subject's level of
education. The probability may vary by education but not by income within that level of education. If the
probability that income is recorded also varies by the value of income within each level of education (for
example, people with high incomes don't report them), then the data are neither MCAR nor MAR. This is
not an uncommon situation, and, if it applies, none of the methods is appropriate.
Related procedures. Many procedures allow you to use listwise or pairwise estimation. Linear Regression
and Factor Analysis allow replacement of missing values by the mean values. In the Forecasting add-on
module, several methods are available to replace missing values in time series.
Missing values 3
Sort by
Cases are listed according to the ascending or descending order of the values of the
specified variable. Available only for All cases, optionally sorted by selected variable.
Listwise method
This method uses only complete cases. If any of the analysis variables have missing values, the case is
omitted from the computations.
Pairwise method
This method looks at pairs of analysis variables and uses a case only if it has nonmissing values for both
of the variables. Frequencies, means, and standard deviations are computed separately for each pair.
Because other missing values in the case are ignored, correlations and covariances for two variables do
not depend on values missing in any other variables.
EM method
This method assumes a distribution for the partially missing data and bases inferences on the likelihood
under that distribution. Each iteration consists of an E step and an M step. The E step finds the
conditional expectation of the "missing" data, given the observed values and current estimates of the
parameters. These expectations are then substituted for the "missing" data. In the M step, maximum
likelihood estimates of the parameters are computed as though the missing data had been filled in.
"Missing" is enclosed in quotation marks because the missing values are not being directly filled in.
Instead, functions of them are used in the log-likelihood.
Roderick J. A. Little's chi-square statistic for testing whether values are missing completely at random
(MCAR) is printed as a footnote to the EM matrices. For this test, the null hypothesis is that the data are
missing completely at random, and the p value is significant at the 0.05 level. If the value is less than
0.05, the data are not missing completely at random. The data may be missing at random (MAR) or not
missing at random (NMAR). You cannot assume one or the other and need to analyze the data to
determine how the data are missing.
Regression method
This method computes multiple linear regression estimates and has options for augmenting the estimates
with random components. To each predicted value, the procedure can add a residual from a randomly
selected complete case, a random normal deviate, or a random deviate (scaled by the square root of the
residual mean square) from the t distribution.
EM Estimation Options
Using an iterative process, the EM method estimates the means, the covariance matrix, and the
correlation of quantitative (scale) variables with missing values.
Distribution
EM makes inferences based on the likelihood under the specified distribution. By default, a
normal distribution is assumed. If you know that the tails of the distribution are longer than
those of a normal distribution, you can request that the procedure constructs the likelihood
function from a Student's t distribution with n degrees of freedom. The mixed normal distribution
also provides a distribution with longer tails. Specify the ratio of the standard deviations of the
mixed normal distribution and the mixture proportion of the two distributions. The mixed
normal distribution assumes that only the standard deviations of the distributions differ. The
means must be the same.
Missing values 5
Maximum iterations
Sets the maximum number of iterations to estimate the true covariance. The procedure stops
when this number of iterations is reached, even if the estimates have not converged.
Save completed data
You can save a dataset with the imputed values in place of the missing values. Be aware, though,
that covariance-based statistics using the imputed values will underestimate their respective
parameter values. The degree of underestimation is proportional to the number of cases that are
jointly unobserved.
To specify EM options
1. In the main Missing Value Analysis dialog, select the variable(s) for which you want to estimate
missing values using the EM method.
2. Select EM in the Estimation group.
3. To specify predicted and predictor variables, click Variables. See the topic “Predicted and Predictor
Variables” on page 7 for more information.
4. Click EM.
5. Select the EM options you want.
Another consideration arises when using the regression method. In multiple regression, the use of a large
subset of independent variables can produce poorer predicted values than a smaller subset. Therefore, a
variable must achieve an F-to-enter limit of 4.0 to be used. This limit can be changed with syntax.
Multiple Imputation
The purpose of multiple imputation is to generate possible values for missing values, thus creating
several "complete" sets of data. Analytic procedures that work with multiple imputation datasets produce
output for each "complete" dataset, plus pooled output that estimates what the results would have been if
the original dataset had no missing values. These pooled results are generally more accurate than those
provided by single imputation methods.
Missing values 7
v Ordinal. A variable can be treated as ordinal when its values represent categories with some intrinsic
ranking (for example, levels of service satisfaction from highly dissatisfied to highly satisfied).
Examples of ordinal variables include attitude scores representing degree of satisfaction or confidence
and preference rating scores.
v Scale. A variable can be treated as scale (continuous) when its values represent ordered categories with
a meaningful metric, so that distance comparisons between values are appropriate. Examples of scale
variables include age in years and income in thousands of dollars.
The procedure assumes that the appropriate measurement level has been assigned to all variables;
however, you can temporarily change the measurement level for a variable by right-clicking the
variable in the source variable list and selecting a measurement level from the pop-up menu. To
permanently change the level of measurement for a variable,
An icon next to each variable in the variable list identifies the measurement level and data type:
Frequency weights. Frequency (replication) weights are honored by this procedure. Cases with negative
or zero replication weight value are ignored. Noninteger weights are rounded to the nearest integer.
Analysis Weight. Analysis (regression or sampling) weights are incorporated in summaries of missing
values and in fitting imputation models. Cases with a negative or zero analysis weight are excluded.
Complex Samples. The Multiple Imputation procedure does not explicitly handle strata, clusters, or other
complex sampling structures, though it can accept final sampling weights in the form of the analysis
weight variable. Also note that Complex Sampling procedures currently do not automatically analyze
multiply imputed datasets. For a full list of procedures that support pooling, see “Analyzing Multiple
Imputation Data” on page 13.
Missing Values. Both user- and system-missing values are treated as invalid values; that is, both types of
missing values are replaced when values are imputed and both are treated as invalid values of variables
used as predictors in imputation models. User- and system-missing values are also treated as missing in
analyses of missing values.
Replicating results (Impute Missing Data Values). If you want to replicate your imputation results
exactly, use the same initialization value for the random number generator, the same data order, and the
same variable order, in addition to using the same procedure settings.
v Random number generation. The procedure uses random number generation during calculation of
imputed values. To reproduce the same randomized results in the future, use the same initialization
value for the random number generator before each run of the Impute Missing Data Values procedure.
v Case order. Values are imputed in case order.
v Variable order. The fully conditional specification (FCS) imputation method imputes values in the
order specified in the Analysis Variables list.
Analyze Patterns
Analyze Patterns provides descriptive measures of the patterns of missing values in the data, and can be
useful as an exploratory step before imputation.
Optional settings
Analysis Weight
This variable contains analysis (regression or sampling) weights. The procedure incorporates
analysis weights in summaries of missing values. Cases with a negative or zero analysis weight
are excluded.
Output
The following optional output is available:
Summary of missing values
This displays a paneled pie chart that shows the number and percent of analysis
variables, cases, or individual data values that have one or more missing values.
Patterns of missing values
This displays tabulated patterns of missing values. Each pattern corresponds to a group
of cases with the same pattern of incomplete and complete data on analysis variables.
You can use this output to determine whether the monotone imputation method can be
used for your data, or if not, how closely your data approximate a monotone pattern. The
procedure orders analysis variables to reveal or approximate a monotonic pattern. If no
nonmonotone pattern exists after reordering you can conclude that the data have a
monotonic pattern when analysis variables are ordered as such.
Variables with the highest frequency of missing values
This displays a table of analysis variables sorted by percent of missing values in
decreasing order. The table includes descriptive statistics (mean and standard deviation)
for scale variables.
You can control the maximum number of variables to display and minimum percentage
missing for a variable to be included in the display. The set of variables that meet both
criteria are displayed. For example, setting the maximum number of variables to 50 and
the minimum percentage missing to 25 requests that the table display up to 50 variables
that have at least 25% missing values. If there are 60 analysis variables but only 15 have
25% or more missing values, the output includes only 15 variables.
Example. A telecommunications provider wants to better understand service usage patterns in its
customer database. They have complete data for services used by their customers, but the demographic
information collected by the company has a number of missing values. Moreover, these values are not
missing completely at random, so multiple imputation will be used to complete the dataset.
Missing values 9
Analyze > Multiple Imputation > Impute Missing Data Values...
1. Select at least two variables in the imputation model. The procedure imputes multiple values for
missing data for these variables.
2. Specify the number of imputations to compute. By default, this value is 5.
3. Specify a dataset or IBM SPSS Statistics-format data file to which imputed data should be written.
The output dataset consists of the original case data with missing data plus a set of cases with
imputed values for each imputation. For example, if the original dataset has 100 cases and you have
five imputations, the output dataset will have 600 cases. All variables in the input dataset are
included in the output dataset. Dictionary properties (names, labels, etc.) of existing variables are
copied to the new dataset. The file also contains a new variable, Imputation_, a numeric variable that
indicates the imputation (0 for original data, or 1..n for cases having imputed values).
The procedure automatically defines the Imputation_ variable as a split variable when the output
dataset is created. If splits are in effect when the procedure executes, the output dataset includes one
set of imputations for each combination of values of split variables.
Optional settings
Analysis Weight
This variable contains analysis (regression or sampling) weights. The procedure incorporates
analysis weights in regression and classification models used to impute missing values. Analysis
weights are also used in summaries of imputed values; for example, mean, standard deviation,
and standard error. Cases with a negative or zero analysis weight are excluded.
The measurement level alert displays when the measurement level for one or more variables (fields) in
the dataset is unknown. Since measurement level affects the computation of results for this procedure, all
variables must have a defined measurement level.
Scan Data
Reads the data in the active dataset and assigns default measurement level to any fields with a
currently unknown measurement level. If the dataset is large, that may take some time.
Assign Manually
Lists all fields with an unknown measurement level. You can assign measurement level to those
fields. You can also assign measurement level in the Data Editor's Variable List pane.
Since measurement level is important for this procedure, you cannot run this procedure until all fields
have a defined measurement level.
Method
The Method dialog specifies how missing values will be imputed, including the types of models used.
Categorical predictors are indicator (dummy) coded.
Imputation Method
Automatic
Scans the data and uses the monotone method if the data show a monotone pattern of
missing values; otherwise, fully conditional specification is used.
Custom
Select when you are certain of which method you want to use.
Fully conditional specification (MCMC)
This is an iterative Markov chain Monte Carlo (MCMC) method that can be used
when the pattern of missing data is arbitrary (monotone or nonmonotone).
For each iteration and for each variable in the order specified in the variable list,
the fully conditional specification (FCS) method fits a univariate (single
Singularity tolerance. Singular (or non-invertible) matrices have linearly dependent columns, which can
cause serious problems for the estimation algorithm. Even near-singular matrices can lead to poor results,
so the procedure will treat a matrix whose determinant is less than the tolerance as singular. Specify a
positive value.
Constraints
The Constraints dialog allows you to restrict the role of a variable during imputation and restrict the
range of imputed values of a scale variable so that they are plausible. In addition, you can restrict the
analysis to variables with less than a maximum percentage of missing values.
Exclude variables with large amounts of missing data
Typically, analysis variables are imputed and used as predictors without regard to how many
missing values they have, provided they have sufficient data to estimate an imputation model.
You can choose to exclude variables that have a high percentage of missing values. For example,
if you specify 50 as the Maximum percentage missing, analysis variables that have more than
50% missing values are not imputed, nor are they used as predictors in imputation models.
Missing values 11
Output
Display
Controls display of output. An overall imputation summary is always displayed, which includes
tables relating the imputation specifications, iterations (for fully conditional specification method),
dependent variables imputed, dependent variables excluded from imputation, and imputation
sequence. If specified, constraints for analysis variables are also shown.
Imputation model
This displays the imputation model for dependent variables and predictors, and includes
univariate model type, model effects, and number of values imputed.
Descriptive statistics for variables with imputed values
This displays descriptive statistics for dependent variables for which values are imputed.
For scale variables the descriptive statistics include mean, count, standard deviation, min,
and max for the original input data (prior to imputation), imputed values (by
imputation), and complete data (original and imputed values together—by imputation).
For categorical variables the descriptive statistics include count and percent by category
for the original input data (prior to imputation), imputed values (by imputation), and
complete data (original and imputed values together—by imputation).
Iteration History
When the fully conditional specification imputation method is used, you can request a dataset
that contains iteration history data for FCS imputation. The dataset contains means and standard
deviations by iteration and imputation for each scale dependent variable for which values are
imputed. You can plot the data to help assess model convergence.
When you open a dataset, the presence of Imputation_ identifies the dataset as a possible MI dataset.
You can distinguish imputed values from observed values by cell background color, the font, and bold
type (for imputed values). When you create a new dataset in the current session with Impute Missing
Values, markings are turned on by default. When you open a saved data file that includes imputations,
markings are turned off.
Alternatively, you can turn on markings by clicking the imputation marking button at the right edge of
the edit bar in Data View of the Data Editor.
Alternatively, you can select the imputation from the drop-down list in the edit bar in Data View of the
Data Editor.
Relative case position is preserved when selecting imputations. For example, if there are 1000 cases in the
original dataset, case 1034, the 34th case in the first imputation, displays at the top of the grid. If you
select imputation 2 in the dropdown, case 2034, the 34th case in imputation 2, would display at the top of
the grid. If you select Original data in the dropdown, case 34 would display at the top of the grid.
Column position is also preserved when navigating between imputations, so that it is easy to compare
values between imputations.
Sometimes you will need to perform transformations on imputed data. For example, you may want to
take the log of all values of a salary variable and save the result in a new variable. A value computed
using imputed data will be treated as imputed if it differs from the value computed using the original
data.
If you edit an imputed value in a cell of the Data Editor, that cell is still treated as imputed. It is not
recommended to edit imputed values in this way.
Both tabular output and model PMML can be pooled. There is no new procedure for requesting pooled
output; instead, a new tab on the Options dialog gives you global control over multiple imputation
output.
v Pooling of Tabular Output. By default, when you run a supported procedure on a multiple
imputation (MI) dataset, results are automatically produced for each imputation, the original
(unimputed) data, and pooled (final) results that take into account variation across imputations. The
statistics that are pooled vary by procedure.
Missing values 13
v Pooling of PMML. You can also obtain pooled PMML from supported procedures that export PMML.
Pooled PMML is requested in the same way as, and is saved instead of, non-pooled PMML.
Unsupported procedures produce neither pooled output nor pooled PMML files.
Levels of pooling
Coefficients (regression and correlation), means (and mean differences), and counts are typically pooled.
When the standard error of the statistic is available, then univariate pooling is used; otherwise naïve
pooling is used.
The following procedures support MI datasets, at the levels of pooling specified for each piece of output.
Generalized Linear Models and Generalized Estimating Equations. These procedures support pooled
PMML.
v The Categorical Variable Information table supports N and Percents at Naïve pooling.
v The Continuous Variable Information table supports N and Mean at Naïve pooling.
v The Parameter Estimates table supports the coefficient, B, at Univariate pooling.
v The Estimated Marginal Means: Estimation Coefficients table supports Mean at Naïve pooling.
v The Estimated Marginal Means: Estimates table supports Mean at Univariate pooling.
v The Estimated Marginal Means: Pairwise Comparisons table supports Mean Difference at Univariate
pooling.
Missing values 15
Multinomial Logistic Regression. This procedure supports pooled PMML.
v The Parameter Estimates table supports the coefficient, B, at Univariate pooling.
Tests for Several Independent Samples. The following features are supported:
v The Ranks table supports Mean Rank and N at Naïve pooling.
v The Frequencies table supports Counts at Naïve pooling.
Tests for Several Related Samples. The following features are supported:
v The Ranks table supports Mean Rank at Naïve pooling.
Analysis Output. This group controls the type of Viewer output produced whenever a multiply imputed
dataset is analyzed. By default, output will be produced for the original (pre-imputation) dataset and for
each of the imputed datasets. In addition, for those procedures that support pooling of imputed data,
final pooled results will be generated. When univariate pooling is performed, pooling diagnostics will
also display. However, you can suppress any output you do not want to see.
Missing values 17
18 IBM SPSS Missing Values 26
Notices
This information was developed for products and services offered in the US. This material might be
available from IBM in other languages. However, you may be required to own a copy of the product or
product version in that language in order to access it.
IBM may not offer the products, services, or features discussed in this document in other countries.
Consult your local IBM representative for information on the products and services currently available in
your area. Any reference to an IBM product, program, or service is not intended to state or imply that
only that IBM product, program, or service may be used. Any functionally equivalent product, program,
or service that does not infringe any IBM intellectual property right may be used instead. However, it is
the user's responsibility to evaluate and verify the operation of any non-IBM product, program, or
service.
IBM may have patents or pending patent applications covering subject matter described in this
document. The furnishing of this document does not grant you any license to these patents. You can send
license inquiries, in writing, to:
For license inquiries regarding double-byte (DBCS) information, contact the IBM Intellectual Property
Department in your country or send inquiries, in writing, to:
This information could include technical inaccuracies or typographical errors. Changes are periodically
made to the information herein; these changes will be incorporated in new editions of the publication.
IBM may make improvements and/or changes in the product(s) and/or the program(s) described in this
publication at any time without notice.
Any references in this information to non-IBM websites are provided for convenience only and do not in
any manner serve as an endorsement of those websites. The materials at those websites are not part of
the materials for this IBM product and use of those websites is at your own risk.
IBM may use or distribute any of the information you provide in any way it believes appropriate without
incurring any obligation to you.
19
Licensees of this program who wish to have information about it for the purpose of enabling: (i) the
exchange of information between independently created programs and other programs (including this
one) and (ii) the mutual use of the information which has been exchanged, should contact:
Such information may be available, subject to appropriate terms and conditions, including in some cases,
payment of a fee.
The licensed program described in this document and all licensed material available for it are provided
by IBM under terms of the IBM Customer Agreement, IBM International Program License Agreement or
any equivalent agreement between us.
The performance data and client examples cited are presented for illustrative purposes only. Actual
performance results may vary depending on specific configurations and operating conditions.
Information concerning non-IBM products was obtained from the suppliers of those products, their
published announcements or other publicly available sources. IBM has not tested those products and
cannot confirm the accuracy of performance, compatibility or any other claims related to
non-IBMproducts. Questions on the capabilities of non-IBM products should be addressed to the
suppliers of those products.
Statements regarding IBM's future direction or intent are subject to change or withdrawal without notice,
and represent goals and objectives only.
This information contains examples of data and reports used in daily business operations. To illustrate
them as completely as possible, the examples include the names of individuals, companies, brands, and
products. All of these names are fictitious and any similarity to actual people or business enterprises is
entirely coincidental.
COPYRIGHT LICENSE:
This information contains sample application programs in source language, which illustrate programming
techniques on various operating platforms. You may copy, modify, and distribute these sample programs
in any form without payment to IBM, for the purposes of developing, using, marketing or distributing
application programs conforming to the application programming interface for the operating platform for
which the sample programs are written. These examples have not been thoroughly tested under all
conditions. IBM, therefore, cannot guarantee or imply reliability, serviceability, or function of these
programs. The sample programs are provided "AS IS", without warranty of any kind. IBM shall not be
liable for any damages arising out of your use of the sample programs.
Each copy or any portion of these sample programs or any derivative work, must include a copyright
notice as follows:
© IBM 2019. Portions of this code are derived from IBM Corp. Sample Programs.
Adobe, the Adobe logo, PostScript, and the PostScript logo are either registered trademarks or trademarks
of Adobe Systems Incorporated in the United States, and/or other countries.
Intel, Intel logo, Intel Inside, Intel Inside logo, Intel Centrino, Intel Centrino logo, Celeron, Intel Xeon,
Intel SpeedStep, Itanium, and Pentium are trademarks or registered trademarks of Intel Corporation or its
subsidiaries in the United States and other countries.
Linux is a registered trademark of Linus Torvalds in the United States, other countries, or both.
Microsoft, Windows, Windows NT, and the Windows logo are trademarks of Microsoft Corporation in the
United States, other countries, or both.
UNIX is a registered trademark of The Open Group in the United States and other countries.
Java and all Java-based trademarks and logos are trademarks or registered trademarks of Oracle and/or
its affiliates.
Notices 21
22 IBM SPSS Missing Values 26
Index
A Missing Value Analysis (continued)
estimating statistics 5
Analyze Patterns 8 expectation-maximization 7
imputing missing values 5
MCAR test 5
C methods 5
correlations patterns 3
in Missing Value Analysis 5, 6 regression 6
covariance missing values
in Missing Value Analysis 5, 6 univariate statistics 4
monotone imputation
in Multiple Imputation 10
E multiple imputation 7
analyze patterns 8
EM impute missing data values 9
in Missing Value Analysis 5 Multiple Imputation 12, 13
extreme value counts
in Missing Value Analysis 4
N
F normal variates
in Missing Value Analysis 6
frequency tables
in Missing Value Analysis 4
fully conditional specification
in Multiple Imputation 10 P
pairwise deletion
in Missing Value Analysis 1
I
Impute Missing Data Values 9
constraints 11 R
imputation method 10 regression
output 12 in Missing Value Analysis 6
incomplete data residuals
see Missing Value Analysis 1 in Missing Value Analysis 6
indicator variables
in Missing Value Analysis 4
iteration history S
in Multiple Imputation 12 sorting cases
in Missing Value Analysis 3
standard deviation
L in Missing Value Analysis 4
listwise deletion Student's t test
in Missing Value Analysis 1 in Missing Value Analysis 6
Little's MCAR test 5
in Missing Value Analysis 1
T
t test
M in Missing Value Analysis 4
MCAR test tabulating cases
in Missing Value Analysis 1 in Missing Value Analysis 3
mean tabulating categories
in Missing Value Analysis 4, 5, 6 in Missing Value Analysis 4
mismatch
in Missing Value Analysis 4
missing indicator variables
in Missing Value Analysis 4
Missing Value Analysis 1
command additional features 7
descriptive statistics 4
EM 5
23
24 IBM SPSS Missing Values 26
IBM®
Printed in USA