SPSS Statistics For Dummies book cover

SPSS Statistics For Dummies

Authors:
Keith McCormick ,
Jesus Salcedo
Published: September 9, 2020

Overview

The fun and friendly guide to mastering IBM’s Statistical Package for the Social Sciences

Written by an author team with a combined 55 years of experience using SPSS, this updated guide takes the guesswork out of the subject and helps you get the most out of using the leader in predictive analysis.

Covering the latest release and updates to SPSS 27.0, and including more than 150 pages of basic statistical theory, it helps you understand the mechanics behind the calculations, perform predictive analysis, produce informative graphs, and more. You’ll even dabble in programming as you expand SPSS functionality to suit your specific needs. 

  • Master the fundamental mechanics of SPSS
  • Learn how to get data into and out of the program
  • Graph and analyze your data more accurately and efficiently
  • Program SPSS with Command Syntax 

Get ready to start handling data like a pro—with step-by-step instruction and expert advice!

 

The fun and friendly guide to mastering IBM’s Statistical Package for the Social Sciences

Written by an author team with a combined 55 years of experience using SPSS, this updated guide takes the guesswork out of the subject and helps you get the most out of using the leader in predictive analysis.

Covering the latest release and updates to SPSS 27.0, and including more than 150 pages of basic statistical theory, it helps you understand the mechanics behind the calculations, perform predictive

analysis, produce informative graphs, and more. You’ll even dabble in programming as you expand SPSS functionality to suit your specific needs. 

  • Master the fundamental mechanics of SPSS
  • Learn how to get data into and out of the program
  • Graph and analyze your data more accurately and efficiently
  • Program SPSS with Command Syntax 

Get ready to start handling data like a pro—with step-by-step instruction and expert advice!

 

SPSS Statistics For Dummies Cheat Sheet

IBM SPSS Statistics is an application that performs statistical analysis on data. To perform statistical analyses correctly, you need to know the level of measurement of the variables because it defines which summary statistics and graphs should be used. It also helps to know the most commonly used procedures in the Analyze menu and possible conclusions that you can reach after conducting a statistical test. [caption id="attachment_272504" align="alignnone" width="556"] SPSS Statistics Analyze menu[/caption]

Articles From The Book

9 results

Other Software Articles

How to Run an Analysis in SPSS Statistics

After you bring data into SPSS Statistics, the next step is to select a procedure. The Analyze menu contains a list of reporting and statistical analysis categories. Most of the categories are followed by an arrow, which indicates that several analytical procedures are available in the category; these appear on a submenu when the category is selected. To select a procedure, choose Analyze, an analysis category, and then the procedure. The procedure dialog will open. Most data files contain many variables and it's not always easy to remember the properties of each one. You may want to produce documentation, often referred to as a codebook, listing all the information about the variables in the data. SPSS provides the Codebook procedure for viewing variable attributes and reporting summary descriptive tables for each variable. To create a Codebook, choose Analyze→Reports→Codebook, as shown. The following figure shows the Codebook dialog. You’ll need to select the variables of interest and then run the analysis from the procedure dialog. Most procedure dialogs have the same basic components and contain a number of common features. Each procedure dialog contains the following components:

  • Source variables are variables available for the procedure.
  • Target variables are variables used in the procedure. You’ll need to move the source variable(s) to the target variables box
  • Control buttons run, reset, or cancel the procedure.
  • Dialog tabs or buttons control optional specifications.
In the source and target variable lists, the variable label is shown, followed by the variable name in square brackets. If a variable doesn't have a label, only the variable name appears.

You can resize any SPSS dialog. If you make it larger, it's easier to see the variable list. In addition, right-click any variable in the source list to display a description of that variable. And if you are having trouble finding a variable in the source list, in most dialogs, you can type the first letter of the label to display matching variable labels. Repeatedly typing the letter will allow you to move through the list to each variable label beginning with that letter. If you're a fast typist, you can include multiple letters to better narrow your search for variables.

The icons displayed next to variables in the dialog provide information about the variable type and measurement level.

Because SPSS procedures provide a great deal of flexibility, the dialog often can't display all possible choices. The main dialog contains the minimum information required to run the procedure. You can make additional optional specifications in subdialogs. The subdialogs are accessed from the buttons located on the right side of the main dialog or tabs at the top of the dialog.

The name of subdialog if often similar to the name of the equivalent subcommands in SPSS Syntax.

Instead of an OK button, subdialogs have a Continue button, to return to the main dialog. The control buttons that appear along the bottom of the dialog instruct SPSS to perform an action:
  • OK runs the procedure. The OK button is disabled (appears dimmed) until the minimum dialog requirements are completed.
  • Reset resets all specifications made in the dialog and associated subdialogs and keeps the dialog open.
  • Cancel cancels the selections and closes the dialog without running the procedure.
  • Help opens the SPSS Help facility with help relevant to the current dialog.
  • Paste: Pastes SPSS syntax for commands into the Syntax Editor window.
In the Codebook procedure, you’ll need to select the variables to display. You can run the codebook on selected variables or on all variables in the file.
  1. In the Variables box, click the first variable, hold down the shift key, and click the last variable.
  2. Click the arrow to move all the variables to the Codebook Variables box, as shown.
  3. Click OK to run the analysis.
After you move the variables (Step 2), you can make selections on the Output and Statistics tabs. Optionally on the Output tab, you can select variable attributes to display in each table and the order of the tables. By default, all variable attributes are displayed and the tables are in the order shown in the Codebook Variables list. On the Statistics tab, you can select statistics to display in the tables. By default, counts and percentages are displayed for variables defined as nominal or ordinal measurement level. For scale variables, the mean, standard deviation, and quartiles are displayed.

Other Software Articles

Modules You Can Add to SPSS

IBM SPSS Statistics comes in the form of a base system, but you can acquire additional modules to add to that system. SPSS is available in various licensing editions: the campus editions, subscription plans, and commercial editions. Although the pricing and various bundles differ for each, they all enable you to include the same add-on modules. If you're using a copy of SPSS at work or in a university setting that someone else installed, you might have some of these add-ons without realizing it because most are so fully integrated into the menus that they look like integral parts of the base system. If you notice that your menus are shorter or longer than someone else’s copy of SPSS, this is probably due to add-on modules. Some add-ons might be of no interest to you; while others could become indispensable. Note that if you have a trial copy of SPSS, it likely has all the modules, including those that you might lose access to when you acquire your own copy. This article introduces you to the modules that can be added to SPSS and what they do; refer to the documentation that comes with each module for a full tutorial.

You'll likely come across the names IBM SPSS Amos and IBM SPSS Modeler. Although SPSS appears in the names, you purchase these programs separately, not as add-ons. Amos is used for Structural Equation Modeling (SEM) and SPSS Modeler is a predictive analytics and machine learning workbench.

The Advanced Statistics module

Following is a list of the statistical techniques that are part of the Advanced Statistics module:
  • General linear models (GLM)
  • Generalized linear models (GENLIN)
  • Linear mixed models
  • Generalized estimating equations (GEE) procedures
  • Generalized linear mixed models (GLMM)
  • Survival analysis procedures
Although these procedures are among the most advanced in SPSS, some are quite popular. For instance, hierarchical linear modeling (HLM), part of linear mixed models, is common in educational research. HLM models are statistical models in which parameters vary at more than one level. For instance, you may have data that includes information for both students and schools, and in an HLM model you can simultaneously incorporate information from both levels. The key point is that this Advanced Statistical module contains specialized techniques that you need to use if you don’t meet the assumptions of plain-vanilla regression and analysis of variance (ANOVA). These techniques are more of an ANOVA flavor. Survival analysis is so-called time-to-event modeling, such as estimating time to death after diagnosis.

The Custom Tables module

The Custom Tables module has been the most popular module for years, and for good reason. If you need to squeeze a lot of information into a report, you need this module. For instance, if you do survey research and want to report on the entire survey in tabular form, the Custom Tables module can come to your rescue because it allows you to easily present vast information.

Get a free trial copy of SPSS Statistics with all the modules, and force yourself to spend a solid day using the modules you don’t have. See if any aspect of reporting you’re already doing could be done faster with the Custom Tables module. Reproduce a recent report, and see how much time you might save.

In the following figure, you see a simple Frequency table displaying two variables. Note that the categories for both variables are the same. The following table is the same data, but here the table was created using the SPSS Custom Tables module and is a much better table. If you’re producing the table for yourself, presentation may not matter. But if you’re putting the table in a report that will be sent to others, you need the SPSS Custom Tables module. By the way, with practice, it takes only a few seconds to make the custom version, and you can use Syntax to further customize the table!

Starting in version 27, the Custom Tables module is part of the standard edition.

The Regression module

The following is a list of the statistical techniques that are part of the Regression module:
  • Multinomial and binary logistic regression
  • Nonlinear regression (NLR) and constrained nonlinear regression (CNLR)
  • Weighted least squares regression and two-stage least squares regression
  • Probit analysis
In some ways, the Regression module is like the Advanced Statistics module — you use these techniques when you don’t meet the standard assumptions. However, with the Regression module, the techniques are fancy variants of regression when you can’t do ordinary least squares regression. Binary logistic regression is popular and used when the dependent variable has two categories — for example, stay or go (churn), buy or not buy, or get a disease or not get a disease.

The Categories module

The Categories module enables you to reveal relationships among your categorical data. To help you understand your data, the Categories module uses perceptual mapping, optimal scaling, preference scaling, and dimension reduction. Using these techniques, you can visually interpret the relationships among your rows and columns. The Categories module performs its analysis on ordinal and nominal data. It uses procedures similar to conventional regression, principal components, and canonical correlation. It performs regression using nominal or ordinal categorical predictor or outcome variables. The procedures of the Categories module make it possible to perform statistical operations on categorical data:
  • Using the scaling procedures, you can assign units of measurement and zero-points to your categorical data, which gives you access to new groups of statistical functions because you can analyze variables using mixed measurement levels.
  • Using correspondence analysis, you can numerically evaluate similarities among nominal variables and summarize your data according to components you select.
  • Using nonlinear canonical correlation analysis, you can collect variables of different measurement levels into sets of their own, and then analyze the sets.
You can use this module to produce a couple of useful tools:
  • Perceptual map: A high-resolution summary chart that serves as a graphic display of similar variables or categories. A perceptual map gives you insights into relationships among more than two categorical variables.
  • Biplot: A summary chart that makes it possible to look at the relationships among products, customers, and demographic characteristics.

The Data Preparation module

Let’s face it: Data preparation is no fun. We’ll take all the help we can get. No module will eliminate all the work for the human in this human–computer partnership, but the Data Preparation module will eliminate some routine, predictable aspects. This module helps you process rows and columns of data. For rows of data, it helps you identify outliers that might distort your data. As for variables, it helps you identify the best ones, and lets you know that you could improve some by transforming them. It also enables you to create special validation rules to speed up your data checks and avoid a lot of manual work. Finally, it helps you identify patterns in your missing data.

Starting in version 27, the Data Preparation and Bootstrapping modules are part the base edition.

The Decision Trees module

Decision trees are, by far, the most popular and well-known data mining technique. In fact, entire software products are dedicated to this approach. If you aren’t sure whether you need to do data mining but you want to try it out, using the Decision Trees module would be one of the best ways to attempt data mining because you already know your way around SPSS Statistics. The Decision Trees module doesn’t have all the features of the decision trees in SPSS Modeler (an entire software package dedicated to data mining), but there is plenty here to give you a good start. What are decision trees? Well, the idea is that you have something you want to predict (the target variable) and lots of variables that could possibly help you do that, but you don’t know which ones are most important. SPSS indicates which variables are most important and how the variables interact, and helps you predict the target variable in the future. SPSS supports four of the most popular decision tree algorithms: CHAID, Exhaustive CHAID, C&RT, and QUEST.

The Forecasting module

You can use the Forecasting module to rapidly construct expert time-series forecasts. This module includes statistical algorithms for analyzing historical data and predicting trends. You can set it up to analyze hundreds of different time series at once instead of running a separate procedure for each one. The software is designed to handle the special situations that arise in trend analysis. It automatically determines the best-fitting autoregressive integrated moving average (ARIMA) or exponential smoothing model. It automatically tests data for seasonality, intermittency, and missing values. The software detects outliers and prevents them from unduly influencing the results. The generated graphs include confidence intervals and indicate the model’s goodness of fit. As you gain experience at forecasting, the Forecasting module gives you more control over every parameter when you’re building your data model. You can use the expert modeler in the Forecasting module to recommend starting points or to check calculations you’ve done by hand. In addition, an algorithm called Temporal Causal Modeling (TCM) attempts to discover key causal relationships in time-series data by including only inputs that have a causal relationship with the target. This differs from traditional time-series modeling, where you must explicitly specify the predictors for a target series.

The Missing Values module

The Data Preparation module seems to have missing values covered, but the Missing Values module and the Data Preparation module are quite different. The Data Preparation module is about finding data errors; its validation rules will tell you whether a data point just isn’t right. The Missing Values module, on the other hand, is focused on when there is no data value. It attempts to estimate the missing piece of information using other data you do have. This process is called imputation, or replacing values with an educated guess. All kinds of data miners, statisticians, and researchers — especially survey researchers — can benefit from the Missing Values module.

The Bootstrapping module

Hang on tight because we’re going to get a little technical. Bootstrapping is a technique that involves resampling with replacement. The Bootstrapping module chooses a case at random, makes notes about it, replaces it, and chooses another. In this way, it’s possible to choose a case more than once or not at all. The net result is another version of your data that is similar but not identical. If you do this 1,000 times (the default), you can do some powerful things indeed. The Bootstrapping module allows you to build more stable models by overcoming the effect of outliers and other problems in your data. Traditional statistics assumes that your data has a particular distribution, but this technique avoids that assumption. The result is a more accurate sense of what’s going on in the population. Bootstrapping, in a sense, is a simple idea, but because bootstrapping takes a lot of computer horsepower, it’s more popular now than when computers were slower. Bootstrapping is a popular technique outside SPSS as well, so you can find articles on the web about the concept. The Bootstrapping module lets you apply this powerful concept to your data in SPSS Statistics.

The Complex Samples module

Sampling is a big part of statistics. A simple random sample is what we usually think of as a sample — like choosing names out of a hat. The hat is your population, and the scraps of paper you choose belong to your sample. Each slip of paper has an equal chance of being chosen. Research is often more complicated than that. The Complex Sample module is about more complicated forms of sampling: two stage, stratified, and so on. Most often, survey researchers need this module, although many kinds of experimental researchers may benefit from it too. The Complex Samples modules helps you design the data collection, and then takes the design into account when calculating your statistics. Nearly all statistics in SPSS are calculated with the assumption that the data is a simple random sample. Your calculations can be distorted when this assumption is not met.

The Conjoint module

The Conjoint module provides a way for you to determine how each of your product’s attributes affect consumer preference. When you combine conjoint analysis with competitive market product research, it’s easier to zero in on product characteristics that are important to your customers. With this research, you can determine which product attributes your customers care about, which ones they care about most, and how you can do useful studies of pricing and brand equity. And you can do all this before incurring the expense of bringing new products to market.

The Direct Marketing module

The Direct Marketing module is a little different from the others. It’s a bundle of related features in a wizardlike environment. The module is designed to be one-stop shopping for marketers. The main features are recency, frequency, and monetary (RFM) analysis, cluster analysis, and profiling:
  • RFM analysis: RFM analysis reports back to you about how recently, how often, and how much your customers spent on your business. Obviously, customers who are currently active, spend a lot, and spend often, are your best customers.
  • Cluster analysis: Cluster analysis is a way of segmenting your customers into different customer segments. Typically, you use this approach to match different marketing campaigns to different customers. For example, a cruise line may try different covers on the travel catalog going out to customers, with the adventurous types getting Alaska or Norway on the cover, and the umbrella-drink crowd getting pictures of the Caribbean.
  • Profiling: Profiling helps you see which customer characteristics are associated with specific outcomes. In this way, you can calculate the propensity score that a particular customer will respond to a specific campaign. Virtually all these features can be found in other areas of SPSS, but the wizardlike environment of the Direct Marketing module makes it easy for marketing analysts to be able produce useful results when they don’t have extensive training in the statistics behind the techniques.

The Exact Tests module

The Exact Tests module makes it possible to be more accurate in your analysis of small datasets and datasets that contain rare occurrences. It gives you the tools you need to analyze such data conditions with more accuracy than would otherwise be possible. When only a small sample size is available, you can use the Exact Tests module to analyze the smaller sample and have more confidence in the results. Here, the idea is to perform more analyses in a shorter period of time. This module allows you to conduct different surveys rather than spend time gathering samples to enlarge your base of surveys. The processes you use, and the forms of the results, are the same as those in the base SPSS system, but the internal algorithms are tuned to work with smaller datasets. The Exact Tests module provides more than 30 tests covering all the nonparametric and categorical tests you normally use for larger datasets. Included are one-sample, two-sample, and k-sample tests with independent or related samples, goodness-of-fit tests, tests of independence, and measures of association.

The Neural Networks module

A neural net is a latticelike network of neuronlike nodes, set up within SPSS to act something like the neurons in a living brain. The connections between these nodes have associated weights (degrees of relative effect), which are adjustable. When you adjust the weight of a connection, the network is said to learn. In the Neural Network module, a training algorithm iteratively adjusts the weights to closely match the actual relationships among the data. The idea is to minimize errors and maximize accurate predictions. The computational neural network has one layer of neurons for inputs and another for outputs, with one or more hidden layers between them. The neural network can be used with other statistical procedures to provide clearer insight. Using the familiar SPSS interface, you can mine your data for relationships. After selecting a procedure, you specify the dependent variables, which may be any combination of continuous and categorical types. To prepare for processing, you lay out the neural network architecture, including the computational resources you want to apply. To complete preparation, you choose what to do with the output:
  • List the results in tables.
  • Graphically display the results in charts.
  • Place the results in temporary variables in the dataset.
  • Export models in XML-formatted files.

Other Software Articles

10 SPSS Statistics Gotchas

Our 10 gotchas serve as a checklist of potential causes of your SPSS Statistics woes. Some just waste your time, but others can both waste your time and ruin your analysis. This list reinforces the importance of avoiding these common issues so you can efficiently use SPSS. Some of these 10 gotchas can be confusing at first. Others are straightforward, but new users might not attribute to them the importance they deserve. What they all have in common is that ignorance of them can get you into hot water. Whenever something seems to be amiss in SPSS, double-check this list. To earn its way onto this list, these gotchas must have generated hundreds of real-world problems as witnessed by us in our client interactions.

Failing to declare level of measurement

To many new users of SPSS, declaring Level of Measurement seems like a nuisance. You can safely ignore it for a while, but our advice is to not wait until the day that it starts causing problems. Here are just a few noteworthy situations where you will regret a decision to procrastinate getting your datasets set up properly:
  • A variable that you need might not appear in a dialog.
  • Features that rely on metadata, such as Codebook, will produce poor results.
  • The chart dialogs won’t offer you the options you need for a particular variable.
  • The Custom Tables add-on module will behave strangely.
Proper metadata is a must for the efficient use of SPSS. Those who attempt to save time by skipping the step of setting up their datasets properly will never succeed because they'll waste time in the long run trying to figure out why SPSS is not behaving as it should.

Conflating string values with labels

Avoid using the string variable type. Instead, use a combination of values and value labels. Back in the 60s and 70s, RAM and hard drive space were expensive and limited. Strings use many more characters and bytes than numerics, and back then SPSS couldn’t perform calculations using RAM alone, so it needed to use the hard drive as we might use a scratch pad. Now, it might seem quaint to worry about such things, but avoiding strings is still core to the design philosophy of SPSS. So what kinds of variables should be stored as strings? Addresses, open-ended comments in survey data, and the names of people and companies are good examples of string variables. There aren’t many more. The names of the 50 states, the names of products, product categories and SKUs, and most other nominal variables should be set up as pairs of values and value labels. In the past, leading zeros in data such as zip codes posed a problem, so the data would be declared as string. Now, however, the restricted numeric variable type adds leading zeros padded to the maximum width of the variable, so a zip code variable no longer needs to be declared as a string. Also, Autorecode makes conversions from string to numeric easy. Keep string variables to a minimum.

Excel files do not allow for metadata, so Excel does not support value and value label pairs. When frequently importing string data from Excel, consider learning the syntax commands as well as autorecode transformation because these techniques might be helpful.

Failing to declare missing data

Years ago, an SPSS user in one of our classes experienced the following situation. He had a 1 through 10 scale, with 10 as the highest satisfaction rating and 1 as the lowest satisfaction rating. He needed a code to represent “refused to answer” and chose 11. When he learned about missing data in class, he wondered if just leaving the 11s in the data would be okay because he had already completed the analysis and the number of refusals was fairly low. You bet it caused a big problem! It could move the average satisfaction quite far towards 11 even with a 1 to 2 percent non-response. What was striking about this example was that the most common answer, 1, was very far from the coded-value for non-response. That fact should have made the analysis obviously wrong and easy to spot. Worse, it is well understood in survey research that refusals often reflect respondents who are highly dissatisfied but reluctant to share their opinion. The choice of 11 made their opinion look highly satisfied, not highly dissatisfied, distorting the results even more. Sadly, folks forget to declare missing quite often, and the error often persists through the final steps of the analysis and is never uncovered. In the example, the problem could have been fixed with one simple step: Declare 11 as user-defined missing. Be vigilant about declaring missing data values in your metadata.

Failing to find add-on modules and plug-ins

What can go wrong with add-on modules? The problem that we observe often with clients is that they read about features in add-on modules and then can’t find the modules. This might seem odd. Wouldn't everyone know which SPSS functions they own? But you, too, could be confused for several reasons:
  • Someone else paid for your copy of SPSS, often a copy that you access at school or work
  • The paperwork for your copy of SPSS says Standard or Premium, but it's not clear what this means.
  • You try to find the module in the menus, referring to an image in a book or blog post, and your screen doesn't look like the image.
  • You borrow some working SPSS syntax from a colleague or book, but it fails to work on your copy of SPSS.
SPSS implements add-on modules by adding them to your menus, typically in the Analyze main menu. In the following figure, you can see the Analyze menu from the screen of an SPSS Subscription trial. The trial version always has all modules. So, if your menu is shorter than the one you see in the image, you know you don't have the full complement of add-on modules. Nothing is wrong with your copy of SPSS. You just don’t have access to all features, including via SPSS Syntax. Some believe that if you know the necessary code and bypass the graphical user interface, you can run any command, but that is not true. To run the syntax for an add-on module, you must own the module. We stress this point because we have seen people borrow Syntax from a source, colleague, or book, and try to copy and paste the code into the Syntax window. The syntax code will not work if you lack the proper licensing. Another common source of confusion is that many SPSS users don't realize that they have access to add-on modules at work or school. This is unfortunate because the modules can be extremely useful. We always recommend the Custom Tables module to clients for greater efficiency in their analysis. Countless times, clients have thought that they had no modules only to discover that Custom Tables was visible in the menus and functioning. Finally, “plug-ins” are a little different than add-on modules. Features can be added to SPSS by using Python and R. If you're a programmer, you could consider doing this task yourself. However, many of these extensions are already available. All you have to do is download them, and they will appear as additional menu items, with a plus symbol next to the menu entry (see the margin icon). Retired SPSSer Jon Peck was instrumental in adding this programmability feature to SPSS.

Failing to meet statistical and software assumptions

SPSS is not that smart. SPSS will do whatever you ask it to do. So, if you have a variable like Marital Status, with the values: 1= Married, 2=Divorced, 3=Separated, 4=Widowed, and 5=Single, and you ask SPSS to give you a mean for Marital Status, SPSS will give you a mean. However, a mean of 2.33 for a nominal variable like Marital Status is not useful. Similarly, if you analyze your data and find that 100% of your friends that you surveyed think that more monetary resources should be devoted to the tennis center at your country club, but you only interviewed tennis players, then you cannot pass off your results as a random sample of country club members, nor can you be surprised with your findings. It is important that you have reliable and valid data. SPSS assumes that your data comes from a random sample; if this is not the case, you can still obtain descriptive information, however you will not be able to generalize your results to a population. You will also need to know what information you can glean from your data. Additionally, it is important to remember that every statistical test has assumptions. Some statistical tests in SPSS, like the independent samples t-test, automatically assess some of the test assumptions, however most of the time; you will have to run additional checks to assess test assumptions. Remember that the better you meet test assumptions, the more you can trust the results of a test.

You may hear that a test is sensitive to violations of assumptions or robust to violations of assumptions. When a test is sensitive, you have to be especially careful to meet the assumptions. When a test is robust, there is more wiggle room with the assumptions.

Confusing fasting syntax with copy and paste

Virtually all SPSS users start by learning SPSS via the Graphical User Interface and many find SPSS Syntax to be a bit arcane. The confusion arises when a colleague shares a bit of syntax code and offers it up as a shortcut, but it can all look very intimidating. The fear is that you will have to have a big book open on your desk and that you will be typing the commands letter by letter. This is simply not true. Even if a well-meaning colleague exclaims “It’s easy, just paste it,” it might not be clear what they mean. “Pasting” in SPSS, in regards to SPSS Syntax, means to let the SPSS dialogs generate the syntax code for you by giving the instructions via point and click. The syntax is then generated and sent to the Syntax Window. You can think of it as converting clicks into code. It is not the copy, paste maneuver (Control-C, Control-V in Windows) that we do in most software.

Thinking you create variables in SPSS as you Do in Excel

Almost everyone who learns SPSS brings prior exposure to Excel to the learning experience. There is a critical function in both which is handled quite differently in the two interfaces. In Excel, when you want to implement a formula you work directly in a cell of the spreadsheet and the formula is saved in that same location when you save the spreadsheet. In SPSS, you must use the Compute Variable dialog (or the equivalent in SPSS Syntax) and your formula is not saved in the dataset @@md only the result is saved in the dataset. At first, it might seem highly desirable for everyone to save formulas in the dataset, but it might not be clear the high price that is paid for this feature in Excel. SPSS is built to be scalable to large datasets, sometimes 100s of millions of rows of data. In Excel, the spreadsheet must be constantly scanned to update the values of formulas. That scanning, passively and automatically in the background, consumes resources and makes Excel less scalable to very large datasets. Excel becomes noticeably sluggish when datasets are very large for this reason, but Excel was never designed for huge datasets. In SPSS, the data remains constant unless an action prompts a change. To force calculations to update, either the menus must be used again or SPSS Syntax must be run again. Each system is designed with its primary audience in mind. If you are more familiar with how Excel automatically updates calculations, how should you acclimate to SPSS? If most of your data is read in from a file and you proceed directly to analysis then you will probably be quite content using the Graphical User Interface. If you have very large files or if you have a large number of calculations that are made after the data is read in from a file, you will need to learn SPSS Syntax to be productive. By saving those calculations, perhaps dozens or hundreds of them, in the form of SPSS Syntax you can rerun them all quite easily.

Excel currently has a limit of 1,000,000 rows of data, but just a few years ago the limit was much smaller. This is rarely an issue for Excel users as that many rows is usually sufficient. Excel experts can often find a way around this limit, but it is rarely necessary. The technical reason for this limit is that the entire spreadsheet must be accessible to a computer’s memory. SPSS does not require the entire dataset to fit in the computer’s memory. This is important to many SPSS users because thousands of companies with datasets larger than the million-row limit need to analyze their large datasets in SPSS. The IRS is a notable example of an organization that uses SPSS that has datasets much larger than the million-row limit.

Getting confused by listwise deletion

Missing data has often been treated as a chapter-length (or even book-length) topic, but a discussion of that length is not possible in this article. You can handle missing data in many ways, one of which is to use listwise deletion. And being familiar with the term listwise deletion may alert you to what would otherwise seem like strange behavior in SPSS. Imagine that you have a large dataset, with thousands of rows. But when you run a multivariate analysis, SPSS behaves as if you have no data at all. You check the steps multiple times, but all you see in the results are messages that indicate that you have “no valid cases.” What could be happening? Listwise deletion is one method for determining which cases in the dataset are used by SPSS for multivariate analysis. When this method is applied, only cases that are valid for all variables in the analysis are used. Missing just a single cell of information in the case row will cause the entire case to be removed. Why is this common? Imagine that you're collating data on airline passengers. One column records if a passenger chose to purchase an inflight meal, which applies to only coach passengers. Another column records which of two meal choices the person chose during the first-class meal, which applies to only first class passengers. Every row in the dataset will be missing one or the other, resulting in zero rows of data being presented to the multivariate analysis. This situation is common. This short discussion is not sufficient to weigh the pros and cons of using listwise deletion. However, you will now be aware of it when you run into the problem of zero cases being analyzed. Also be on the lookout for times when many fewer cases than you were expecting are analyzed. In the Options dialog of the Linear Regression dialog, listwise deletion is the default. Be careful not to haphazardly choose among the other choices until the regression works. Instead, understand the other options before you try them.

Losing track of your active dataset

Your SPSS skills are progressing along nicely and you decide that it's time to try SPSS Syntax. You double-check your work, run the syntax, and encounter the warning shown here. You confirm that you have the necessary dataset and the necessary variable. What has happened? Almost certainly, you have two (or more) datasets open and you’ve lost track of which one is active. When you're working in the graphical user interface, it's virtually impossible to get confused because when you access the menus and dialogs you're generally doing so from the Data Editor window. When you're using SPSS Syntax, however, you're running code and there's no guarantee that the necessary data elements are present. Here's what you need to do: Check to see if you have more than one dataset open, and ensure that the dataset you need is the active dataset. The Syntax window has the following indicator: DataSet1 is simply the dataset you opened first. To switch to DataSet2, simply click the arrows and select it. You can assign the dataset that you need also by using the following bit of syntax: DATASET ACTIVATE DataSet1.

Forgetting to turn off Select and Split and Weight

A common mistake occurs when you're dealing with a command that stays in effect until you explicitly instruct SPSS to turn it off. Three of these commands are Select, Split, and Weight, which are somewhat unusual in SPSS because they're typically associated with a temporary adjustment to an analysis, not with a permanent change to the data. Weight is more technical and is more often associated with survey analysis. Here is a quick explanation of each:
  • Select: Indicates which cases you want to include or exclude from your analysis
  • Split: Separates the dataset by a grouping variable and analyzes each group separately
  • Weight: Adjusts underrepresented groups as if they were fully represented, and applies the reverse adjustment to overrepresented groups.
Effective use of all three requires more than just a quick definition. However, checking to see if they're still on is easy, due to an indicator in the lower-right corner of the Data Editor window. The Filter indicator refers to operations in the Select Cases dialog. The Weight and Split By indicators refer to the Weight and Split dialogs, respectively. (Unicode refers to the encoding system used by SPSS, which is typically not temporary, although you can change this in the Edit→Options menu.) If SPSS is behaving strangely and you're not getting the results you expect, check these indicators. To turn an indicator off, return to the dialog where you gave the original instruction.

A common mistake is to accidentally use Select and Split at the same time. (Power users of SPSS might do this intentionally, but only rarely.) In particular, it's never a good idea to use Select and Split on the same variable at the same time. If you do, numerous warnings will appear in the SPSS Output Viewer window.