Skip to main content

​
Overview

Behavioral tracking is a powerful feature of Roundtable Alias that analyzes participants’ typing patterns, keystroke dynamics, and other behavioral signals to identify suspicious activities that may indicate fraudulent responses. By monitoring how participants interact with the survey interface, Alias can detect abnormal behaviors that deviate from genuine human response patterns. This added layer of fraud detection complements the content-based checks and helps ensure the integrity of your survey data.

​
Behavioral Signals Monitored

Alias tracks a range of behavioral signals throughout the survey response process, including:
  1. Typing Speed: Measures the average time between keystrokes and flags responses with abnormally fast or slow typing speeds.
  2. Typing Rhythm: Analyzes the consistency and variability of typing patterns to identify bot-like or scripted behaviors.
  3. Pauses and Hesitations: Detects unusual pauses or hesitations that may indicate copy-pasting, programmatic entry, or disengagement.
  4. Keypress Sequences: Monitors the order and frequency of specific keypresses to identify repetitive or unnatural typing patterns.
  5. Revision Patterns: Tracks how participants edit and revise their responses, flagging abnormal patterns such as lack of revisions or excessive backspacing.
These behavioral signals, when combined with content-based checks, provide a comprehensive assessment of response authenticity and quality.

​
Enabling Behavioral Tracking

To leverage Alias’s behavioral tracking capabilities, you need to include the question_histories parameter in your API requests. This parameter captures the detailed typing data for each question response. To collect question_histories data:
  1. Integrate the Alias JavaScript tracker into your survey platform or custom survey interface. The tracker captures the necessary keystroke and timing data.
  2. When submitting survey responses to the Alias API, include the question_histories object, which contains the tracked typing data for each question.
Example question_histories data: 
"question_histories": {
  "Q1": [
    [
      {"s":"I","t":0},
      {"s":"I ","t":253},
      {"s":"I l","t":1290},
      {"s":"I li","t":1563},
      {"s":"I lik","t":1817},
      ...
    ]
  ],
  "Q2": [
    [
      {"s":"T","t":0},
      {"s":"Th","t":180},
      {"s":"The","t":450},
      {"s":"The ","t":890},
      ...
    ]
  ]
}
Refer to the JavaScript Tracker documentation for detailed integration instructions.

​
Interpreting Behavioral Tracking Results

When Alias detects suspicious behavioral patterns, it flags the corresponding responses in the checks object of the API response. Example checks object with behavioral flags: 
"checks": {
  "Q1": [
    "Unnatural typing",
    "Response pasted"
  ],
  "Q2": [
    "Programmatic entry"
  ]
}
In this example, the response to Q1 was flagged for unnatural typing patterns and pasting behavior, while Q2 was flagged for programmatic entry. Use these behavioral flags alongside content-based checks to make informed decisions about response validity and data quality.

​
Customizing Behavioral Thresholds

Alias uses default thresholds for detecting abnormal behavioral patterns based on extensive research and data analysis. However, you can customize these thresholds to better suit your specific survey requirements. Contact our support team at [email protected] to discuss custom behavioral tracking configurations and thresholds.

​
Next Steps