Creating a Survey Code Book
📌 A code book is essential for organizing survey results and facilitating future analysis, even when original survey memory fades.
📌 A good code book allows others to recreate the survey design using only the provided data set.
📌 The first step involves identifying the data type for each question: nominal (no rank, e.g., yes/no), ordinal (ranked categories, e.g., fair to excellent), or ratio/interval (continuous numbers, e.g., blood pressure).
📌 "Mark all that apply" questions are treated as separate nominal (yes/no) data points, while open text responses are labeled as string data.

Data Coding and Entry in Excel
📌 Responses must be assigned numerical codes because Excel analyzes numbers better than words (e.g., Brown hair = 1, Black hair = 2).
📌 For "mark all that apply" questions, assign Yes = 1 and No/Blank = 0 so that the total count of '1's indicates the number of conditions marked.
📌 Each row in the Excel data sheet represents a single survey response, requiring a unique ID number as the first column.
📌 Columns should be labeled concisely using abbreviated question numbers (e.g., Q1, Q3\_cis, Q4a).

Variable View and Documentation
📌 A second sheet, labeled Variable View, should be created to document all variables for reproducibility.
📌 The Variable View lists variables (e.g., ID number, Q1, Q2), the full question text (long title), the corresponding data type (nominal, ordinal, ratio, string), and the value codes (what each number represents).
📌 For ratio data types (like blood pressure readings), no value codes are needed as the number itself is the value.
📌 For string (text) variables, the value code section remains blank as the entry is typed words, not numerical codes.

Transforming and Calculated Variables
📌 Calculated variables, such as Mean Arterial Pressure (MAP), can be derived from existing data using a specific formula.
📌 The formula provided for MAP is: $$\text{MAP} = \frac{\text{Systolic} + (2 \times \text{Diastolic})}{3}$$
📌 Once the formula is entered for the first response in Excel, it can be dragged down to automatically calculate MAP for all subsequent survey responses.

Key Points & Insights
➡️ Establish clear numerical codes for categorical responses (nominal/ordinal) before data entry to enable effective quantitative analysis.
➡️ Utilize a separate Variable View sheet to fully document variable names, full questions, data types, and all associated value codes to ensure survey reproducibility.
➡️ Use 0 for negative/absent answers (e.g., No) and 1 for positive/present answers (e.g., Yes) in dichotomous questions to allow for easy summing of responses.
➡️ Calculate derived metrics like MAP directly within the dataset using formulas to enhance analysis without manual recalculation.

📸 Video summarized with SummaryTube.com on Dec 01, 2025, 06:55 UTC