Preparing Your Data
To ensure smooth processing and the best possible results, please follow these guidelines before importing data.Supported Formats
Supported upload files:- CSV
- Excel
Row and column limits
Each upload file may contain at most 5,000 rows and at most 5,000 columns. If your file exceeds either limit, you need to split or trim the data before you can import it. If the file has too many columns, ReDem will ask you to reduce the file to 5,000 columns maximum (for example by removing variables you do not need for the import) before you can continue. The row limit applies to the full sheet you import (including header rows, question-text rows, and respondent rows). If you need more than 5,000 respondents in one survey, upload a file that fits within the limit first, then add further respondents afterward. See Updating Imported Data for how to append new interviews after the initial upload.Required Formatting for Quality Checks
If you want to use content-based Quality Checks (Coherence Score or Open-Ended Score), your uploaded file must include both questions and answers in a clearly labeled format. You can structure your file in two ways:a) Single-Sheet Files
If all data is stored in one sheet, ensure that each column contains:- A unique variable identifier (e.g.,
Q1A1_1) - The full question text (e.g., “Please indicate how often you usually eat the following types of food. Think about a normal week. / Red meat (e.g., beef, pork)”)
- The labeled answer (e.g., “rarely”)
- Row 1 - Header row: Unique variable identifiers
- Row 2 - Question text row: Full question text for each variable
- Row 3 onward - Respondent data: One row per respondent, with responses in plain text
It is also acceptable if the question text is embedded directly in the header row after the unique variable identifier (e.g.,
Q1A1_1: Please indicate how often you usually eat the following types of food. Think about a normal week. / Red meat (e.g., beef, pork)).b) Two-Sheet Files
If you use separate sheets: Sheet 1 (Data):- Row 1: Unique variable identifiers
- Row 2 onward: Coded answers of the interview
- The mapping between variable identifiers and their full question text
- The labeled answer options for each question and its variable identifier
Quick Import
Follow these steps for a quick import:Upload File
Navigate to Import Data and select the file from your computer. If you want to reuse the configuration from a previous tracking survey, activate Reuse settings from existing survey, select the reference survey, and click Use Template Settings. Otherwise choose “Select Manually” if you want to manually select which questions will be included in the quality check. If your Excel file contains multiple sheets, select which sheet contains the data and which sheet contains the data map.
Provide Metadata
- Assign a survey name.
- Confirm the correct Header Row (usually row 1), Question Text Row (usually row 2), and Respondent Start Row (usually row 3).
- If your dataset uses codes for missing values, enter them.
- Save your settings.
Choose the Identification Column
Select the column that uniquely identifies each respondent (e.g., Respondent ID).
Select Quality Checks and Questions
- Add a New Quality Check: For example, choose Open-Ended Score.
- Choose Data Points: Select the relevant open-ended question.
- Determine Settings: Add the question text, enrich it with keywords if needed, specify the allowed languages (leave empty to allow all languages), and activate or deactivate the duplication check.
- Recommendation: Activate the duplicate check only if identical or highly similar responses to open-ends among and within the interviews are implausible.
- Add Additional Open-Ended Questions: Repeat as necessary, then confirm.
- Configure Other Quality Checks:
- Grid Questions Score: Assign a name to the data point, then select the first and last columns of the grid (minimum of 7 columns required). If the grid is not randomized, enable the Pattern Check to identify fraudulent click patterns beyond simple straightlining.
- Coherence Score: Select all questions you want evaluated for coherence.
- Time Score: Select the variables that capture response time and/or interview length (in seconds or milliseconds).
Determine Cleaning Settings
Decide whether to enable ReDem’s Automatic Data Cleaning. This feature applies exclusion recommendations based on predefined thresholds.
- If Enabled: You can further customize the cleaning settings to suit your project.
- If Disabled: You will still receive the selected quality scores, but no exclusion recommendations will be applied.
- ReDem Score is below 60 OR
- Two or more open-ended questions have an Open-Ended Score below 40 OR
- Two or more answers to open-ended questions are classified as:
- AI-Generated OR
- Gibberish OR
- Off Topic OR
- Wrong Language OR
- Bad Language OR
- Coherence Score is below 30 OR
- Two or more grid questions have a Grid-Question Score below 20 OR
- Time Score is below 30.
Review and Confirm
After configuring your quality checks and cleaning settings, you will see a summary of the selected questions and checks.
- Review your selections carefully.
- Adjust any mistakes before proceeding.
- Check how many credits your chosen checks will consume.
AI Assisted Upload
Upload File
Navigate to Import Data and select the file from your computer. If you want to reuse the configuration from a previous tracking survey, activate Reuse settings from existing survey, select the reference survey, and click Use Template Settings. Otherwise choose “Analyze with AI”. If your Excel file contains multiple sheets, select which sheet contains the data and which sheet contains the data map. Please note that this method may take additional time, as the system will analyze your file, identify the relevant questions for each quality check type, and extract the necessary metadata. Larger files will require more time to process. During this process, you can freely navigate to other pages or even close the application.
Check and confirm the metadata
Check and confirm the extracted metadata for your survey. You can edit each item if needed.
Choose the Identification Column
Select the column that uniquely identifies each respondent (e.g., Respondent ID).
Review AI-Recommended Quality Checks
The AI’s suggestions can require refinement. Manually review each quality score, adjust as needed, and confirm your selections.Open-Ended Score (OES)
- Add keywords (optional), define allowed languages (recommended), and enable or disable the duplicate check.
- Remove any open-ended questions you do not want scored with ReDem’s Open-End Score.
- Confirm your choices.
- The AI identifies grid questions with 7 or more items.
- If any should be excluded - for example, where response patterns like straightlining are acceptable - remove them using the delete icon.
- Confirm your selections.
- Displays which questions are included in the check.
- Ideally, include all questions where possible.
- Select or deselect questions as needed, then confirm.
- AI suggests the variables that capture response time and/or interview length.
- Select or deselect response times/interview length as needed, then confirm.
Determine Cleaning Settings
Decide whether to enable ReDem’s Automatic Data Cleaning. This feature applies exclusion recommendations based on predefined thresholds.
- If Enabled: You can further customize the cleaning settings to suit your project.
- If Disabled: You will still receive the selected quality scores, but no exclusion recommendations will be applied.
- ReDem Score is below 60 OR
- Two or more open-ended questions have an Open-Ended Score below 40 OR
- Two or more answers to open-ended questions are classified as:
- AI-Generated OR
- Gibberish OR
- Off Topic OR
- Wrong Language OR
- Bad Language OR
- Coherence Score is below 30 OR
- Two or more grid questions have a Grid-Question Score below 20 OR
- Time Score is below 30.
Review and Confirm
After configuring your quality checks and cleaning settings, you will see a summary of the selected questions and checks.
- Review your selections carefully.
- Adjust any mistakes before proceeding.
- Check how many credits your chosen checks will consume.
Updating Imported Data
Use this flow to add new interviews to an existing import without reprocessing previously uploaded records.Add data records
Click the green ”+ Add Data Records” button in the top-left to open the Upload Data File screen again.
Upload the updated file
Choose the file that contains additional interviews from the same survey and click “Proceed to Review”.
You may upload all interviews (previous + new) or only the new interviews. Existing interviews are automatically detected and skipped; only new interviews are uploaded.
Check key settings
- Verify the Survey Metadata is correct.
- Review the Respondent ID to ensure proper matching.
- Confirm the Quality Checks are set as intended.
Review summary
Before confirming, a summary will show how many new interviews will be added and the associated credit cost. Only new interviews consume credits.
Workflow for Tracking Surveys
If you import recurring or tracking surveys, the fastest option is to reuse the setup from a previous survey.Reuse an existing setup
After the file upload, activate Reuse settings from existing survey and select the previous survey you want to use as reference.
Apply the reference survey
Click Use Template Settings. ReDem will carry over the reference survey’s metadata, respondent ID and attributes, quality checks, and cleaning settings into the new import.
Edit and Reapply Cleaning Settings
After you have cleaned a file, you can adjust and reapply your cleaning settings at any time.Open Cleaning Settings
On your survey results page, click the “Cleaning Settings” button at the top.
