1. Introduction

Data cleaning and transformation are critical steps in preparing data for analysis. Properly cleaned and transformed data ensures the accuracy and validity of your statistical results. This section covers techniques for identifying and handling missing data, recoding variables, and computing new variables in IBM SPSS Statistics.

2. Identifying and Handling Missing Data

2.1. Identifying Missing Data

1. Descriptive Statistics:

   • Go to Analyze > Descriptive Statistics > Frequencies.
   • Select the variables you want to analyze and click OK.
   • The output will show the number of valid and missing cases for each variable.

2. Missing Value Analysis:

   • Go to Analyze > Missing Value Analysis.
   • Select the variables to include in the analysis.
   • Click OK to generate a report that provides detailed information about the pattern and extent of missing data.

2.2. Handling Missing Data

1. Exclude Cases with Missing Data:

   • Go to Data > Select Cases.
   • Choose If condition is satisfied and click If.
   • Enter a condition to exclude cases with missing values (e.g., NOT MISSING(Age) AND NOT MISSING(Gender)).
   • Click Continue and OK.

2. Replace Missing Values:

   • Go to Transform > Replace Missing Values.
   • Select the variable with missing values and choose a method for replacement (e.g., mean, median, mode).
   • Click OK to replace the missing values.

3. Imputation:

   • Use advanced techniques such as multiple imputation to handle missing data.
   • Go to Analyze > Multiple Imputation > Impute Missing Data Values.
   • Follow the steps to perform multiple imputation.

3. Recoding Variables

3.1. Recode into Same Variables

1. Recode Categorical Variables:
   • Go to Transform > Recode into Same Variables.
   • Select the variables to recode and click Old and New Values.
   • Define the old values and specify the new values (e.g., 1 -> Male, 2 -> Female).
   • Click Continue and OK.

3.2. Recode into Different Variables

1. Recode into New Variables:

   • Go to Transform > Recode into Different Variables.
   • Select the variables to recode and specify the names for the new variables.
   • Click Old and New Values to define the recoding rules.
   • Click Continue and OK.

2. Example:

   • Recode age groups into categories (e.g., 18-25 -> 1, 26-35 -> 2, etc.).

4. Computing New Variables

4.1. Compute Variable

1. Create New Variables:

   • Go to Transform > Compute Variable.
   • Enter the name for the new variable in the Target Variable field.
   • Define the computation formula in the Numeric Expression field (e.g., BMI = Weight / (Height^2)).
   • Click OK to create the new variable.

2. Use Functions:

   • Use built-in functions to perform complex calculations (e.g., MEAN(var1, var2, var3)).

4.2. Example Calculations

1. Sum Scores:

   • Compute the total score from multiple test items (e.g., Total_Score = Score1 + Score2 + Score3).

2. BMI Calculation:

   • Compute Body Mass Index using weight and height variables (e.g., BMI = Weight / (Height^2)).

5. Data Transformation

5.1. Standardizing Variables

1. Z-Score Standardization:
   • Go to Analyze > Descriptive Statistics > Descriptives.
   • Select the variables to standardize and check Save standardized values as variables.
   • Click OK to generate Z-scores for the selected variables.

5.2. Categorizing Continuous Variables

1. Visual Binning:
   • Go to Transform > Visual Binning.
   • Select the continuous variable to bin and click Continue.
   • Define cut points and labels for the new categorical variable.
   • Click OK to create the binned variable.

5.3. Aggregating Data

1. Aggregate Data:
   • Go to Data > Aggregate.
   • Select the variables to group by and the variables to aggregate (e.g., sum, mean).
   • Click OK to generate the aggregated dataset.

Practical Exercises and Datasets for Hands-On Learning

Practical Exercises:

• Exercise 1: Identify and handle missing data in a given dataset using exclusion and replacement techniques.
• Exercise 2: Recode categorical variables and create new variables using computation formulas.
• Exercise 3: Standardize a set of continuous variables and create binned categories.

Datasets:

• Kaggle Datasets - Wide range of datasets for practice.
• UCI Machine Learning Repository - Comprehensive collection of datasets.

Case Studies and Examples from Healthcare Settings

Website Links:

• IBM Case Studies - Healthcare
• PubMed Central - Access numerous case studies and research articles.
• WHO Case Studies

Prototyping Tools and Templates

Website Links:

• Lucidchart - Create visual diagrams and data flow charts.
• Balsamiq - Simple wireframing tool for creating prototypes.
• Microsoft Visio - Advanced diagramming tool.

Recommended Apps and Digital Tools

Website Links:

• RStudio - Integrated development environment for R.
• JASP - Free and user-friendly software for statistical analysis.
• Tableau - Powerful data visualization tool.
• Power BI - Business analytics service by Microsoft.

Supplementary Reading Materials and Resources

Website Links:

• IBM SPSS Statistics Documentation
• Introduction to the Practice of Statistics
• PubMed Central - Access to numerous research articles.
• JAMA Network

End of Topic Quizzes with Answers

Quiz 1: Identifying and Handling Missing Data

1. What is the first step to identify missing data in SPSS?
   • Answer: Analyze > Descriptive Statistics > Frequencies
2. Which method can be used to replace missing values with the mean in SPSS?
   • Answer: Transform > Replace Missing Values

Quiz 2: Recoding and Computing Variables

1. How can you recode a variable into a different variable in SPSS?
   • Answer: Transform > Recode into Different Variables
2. What is the function of Transform > Compute Variable in SPSS?
   • Answer: To create a new variable based on a formula

Relevant Takeaway Assignments

Assignment 1: Use a healthcare dataset to identify and handle missing data. Document the steps and rationale for the methods used.

Assignment 2: Recode age groups into categorical variables and compute a new variable (e.g., BMI) using provided weight and height data.

Assignment 3: Standardize a set of continuous variables and create visualizations to compare the distributions before and after standardization.

These detailed notes, resources, and exercises provide a comprehensive guide to data cleaning and transformation in IBM SPSS Statistics, essential for ensuring data accuracy and reliability in healthcare and medical research.