RESEARCH SUPPORT
SPSS Data Preparation for Research
Variables, Value Labels and Missing Codes
A data file set up so that SPSS reads it correctly: each variable typed and labeled, categories coded with value labels, non-responses declared as missing, and every change logged in syntax before the first analysis is run.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
What Does SPSS Data Preparation Involve?
SPSS data preparation turns a raw spreadsheet or survey export into a file that SPSS interprets correctly. Each variable receives a type, a label and a measurement level, categories receive codes with value labels, and values such as 99 or -1 that stand for no answer are declared user-missing so that they are excluded and not averaged as real data. Derived variables, reversed items and recoded categories are created by documented commands, and the structure of the file is checked against the design, for example one row per participant or one per measurement occasion. The file is prepared, and its coding record kept, by in-house PhD-level statisticians of OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research. Your protocol sets the research questions the file must serve.
WHEN TO ENGAGE
An SPSS File Not Yet Fit for Analysis
A file can open without a warning and still carry coding that produces wrong tables, so the signs usually appear in the first descriptive output.
Spreadsheet Just Imported A survey export or Excel sheet was opened in SPSS, and the variable types, labels and measurement levels were set from the cell contents and have not been checked.
Odd Means and Ranges A mean far outside the plausible range, or an average above the top of a rating scale, often points to missing codes such as 99 being treated as real answers.
Coding Scheme Changed Questionnaire versions or data collectors coded the same answer differently, or two data sources used different codes, and the file needs one consistent scheme before any table is run.
Handover Between Collaborators The file will be used by a supervisor or coauthor, so labels, codes and derived variables must be documented well enough for a stranger to follow.
FILE SET-UP STAGES
From Raw Export to an Analysis-Ready File
Set-up follows an order, because later stages depend on decisions made earlier, such as how missing answers were defined.
Variable Properties
Names, types, labels and measurement levels are set for every variable, so that procedures treat nominal, ordinal and scale variables appropriately.
Variable Coding
SPSS variable coding gives each category a numeric code and a value label, with one consistent scheme across items, waves and data sources, and a codebook that records it.
Missing Value Coding
SPSS missing value coding declares codes such as 97 to 99 as user-missing and keeps refused apart from not applicable, so that later choices, described under Missing Data Analysis, can use the reason.
Recodes and Derived Variables
Reverse-coded items, collapsed categories, age groups and totals are created by documented commands into new variables, leaving the original responses untouched.
Range and Skip Checks
Out-of-range answers, responses that break the questionnaire's skip pattern and answers that contradict each other are found by documented checks before Quantitative Data Analysis begins.
Record of Every Change
Every change is saved in a syntax file and summarized in a codebook, so that the prepared file can be rebuilt from the original data.
SILENT CODING FAULTS
Coding Errors That Survive Into the Results
Preparation errors rarely announce themselves, and every table built on the file inherits them.
Missing Code Inside the Valid Range The code 0 or 9 marks no answer on an item where the same value is also a genuine response, so declaring it missing removes real answers and leaving it counts non-answers.
Multiple Answers in One Cell A check-all-that-apply question arrives as a single text field listing every option chosen, so no option can be counted until each becomes its own coded variable.
Value Labels That Disagree Two waves label code 2 as different answers, and merging them quietly relabels responses, so the same code means different things across the combined file.
Measurement Level Mislabeled A nominal variable is set to scale, or a continuous one to nominal, which changes what dialogs and charts offer and can steer the analysis toward the wrong procedure.
Filters and Splits Left On A Select Cases filter or Split File setting from an earlier session stays active, so later tables describe a subset or appear separately by group without anyone noticing.
Identifiers and Dates as Text Participant identifiers and dates arrive from the spreadsheet as text, so merges fail to match records and date arithmetic is impossible until they are converted.
DELIVERABLES
The Prepared File and Its Coding Record
The file arrives with the record needed to rebuild it, so a later user of the data does not depend on anyone's memory of what was done.
Cleaned SPSS Data File The prepared dataset in SPSS format with typed, labeled and ordered variables, and the original import preserved alongside it.
Codebook A table of variables with labels, codes, value labels, missing value definitions and the source question for each one.
Preparation Syntax The commands that applied every recode, computation and missing value definition, in the order they ran, so the file can be regenerated from the raw export.
Data Issues Log Each problem found, such as a broken skip pattern, an out-of-range answer or a code used two ways, with the decision taken and the number of records affected.
Missing Value Profile The extent of missing data in each variable and the missing value codes in use, giving a first view of how much information each analysis will lose.
Properties for Later Waves The variable properties, value labels and missing value definitions saved so that they can be copied onto a later wave or an extra site, which then arrives coded the same way.
ILLUSTRATIVE EXAMPLE
An Exported Survey Where 99 Meant Two Things
An illustration of how coding choices in a raw export affected the results before any test was run.
ILLUSTRATIVE EXAMPLE · AIRPORT PASSENGER SURVEY RESEARCH · UNITED STATES
A transportation researcher exported an intercept survey of departing passengers from a survey platform and wanted to compare satisfaction with security screening across terminals.
The export read cleanly into SPSS, but the frequency tables showed that the satisfaction items ran from 1 to 5 with 99 for refused and 98 for not applicable, and that the platform had worded one item in reverse. Left alone, the 99 codes would have inflated the mean satisfaction. Both codes were declared user-missing, the reversed item was recoded into a new variable, and the terminal variable, which carried different labels in two export batches, was harmonized and checked against the raw responses. Passengers who agreed to answer were not a random sample, so the pattern of missing responses was summarized and not assumed to be random.
Outcome: The researcher received the prepared SPSS file, the codebook, the preparation syntax, the data issues log and the descriptive tables with 95% confidence intervals for the means.
This example does not describe a real client or project. We never share details of actual client work, in line with the confidentiality we promise every researcher. This illustrative example is included to show the type, scope, and standard of work involved in this kind of analysis.
SOFTWARE
When a Second Program Touches the Export
Raw data come from survey platforms, spreadsheets and other statistical packages, and the preparation has to survive the move between them.
HOW WE ASSESS YOUR PROJECT
We Review Your Actual Research
Your assessment is based on a review of your actual research materials and analytical requirements, not a generic project category.
You Submit Your Project Details
Share whatever is available at the current stage, such as the research aim, protocol or proposal, dataset, variable list, manuscript or statistical requirements.
We Review Your Research and Define the Scope
Your statistically trained Research Coordinator reviews the materials to understand the study design, research questions, data structure, analytical requirements and expected deliverables. If clarification is needed, additional documents or information can be requested before the assessment is finalized.
Your Analysis Is Completed and Delivered
Once the scope and fixed project fee are agreed, the in-house statistician completes the work and the Research Coordinator remains the communication point through delivery, with the analysis, deliverables and timeline as agreed.
RESEARCH RESPONSIBILITY
Coding Choices and the Meaning of Each Variable
A prepared file is a statistical product, and decisions about which cases belong in the study and what the variables mean for the field rest with the researcher.
Agreed Statistical Support OnlineSPSS provides the agreed methodological, analytical and statistical reporting support, defined in a scope that is confirmed before any work begins.
No Intellectual Contribution OnlineSPSS does not claim any intellectual contribution to your study or to the dataset prepared for it. Authorship remains the researcher's.
Your Research Decisions The research question, the definitions of the constructs, the inclusion criteria and the conclusions drawn from the data remain with you, along with overall responsibility.
Confidential Handling Research data and project details are handled confidentially, project information is not published, and an NDA is available where required.
FREQUENTLY ASKED QUESTIONS
Blank Cells, Odd Values and Measurement Levels
How to treat 99, what to do with an impossible value and what the measurement setting changes come up while a file is still being built.
What does SPSS data preparation include for a research project?
SPSS data preparation at OnlineSPSS covers typing and labeling variables, coding categories, defining missing values, creating derived variables and checking the structure of the file, with every change recorded. Paper questionnaires keyed by hand, spreadsheets and survey platform exports are all prepared by the same documented route.
Should missing values be left blank or coded as 99?
Either can work, provided SPSS knows. A blank numeric cell becomes system-missing automatically, whereas a code such as 99 must be declared as a user-missing value in the Missing column of Variable View, otherwise it is read as a real number. Separate codes for refused, not applicable and not asked keep information that blanks lose.
What should be done with an impossible value such as an age of 250?
The source record is checked first, because a keying error can often be corrected from the questionnaire or the collection log. A value that cannot be verified is set to missing by a documented command and not deleted, and the case stays when its other values are sound. The decision, the reason and the number of values affected go in the data issues log.
Does the measurement level matter in SPSS?
Yes, in a limited way. The nominal, ordinal or scale setting affects what some procedures and charts offer by default, but it does not change the numbers in the file. Treating a five-point rating as a scale in a model is a statistical decision about the analysis, and the setting does not make that decision.
How do you protect my research data and confidentiality?
Your data and project details are handled confidentially, and project information is not published. Researcher names and institutions are not used publicly without permission. An NDA is available where required and can be arranged before sensitive materials are shared. Read more about confidentiality and data security.
What happens when I request an assessment, and how long does it take?
You share your research question, design and relevant materials, such as the raw export, the questionnaire and any codebook you already have, and a statistically trained Research Coordinator reviews the request and responds the same day. The analysis, deliverables, timeline and one fixed project fee are then confirmed before any work begins, so the timeline follows from the agreed scope.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
A Data File
That Explains Itself
Missing value definitions and value labels travel inside the .sav file, so a dataset prepared once carries its own meaning to every procedure and every coauthor who opens it.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014