Original studies & reports
Peer-reviewed research, industry reports, and whitepapers from Tayqun and the wider research community across MENA.
The State of Online Survey Quality in 2026
An analysis of 1.24 million responses across 4,331 studies: how prevalent low-quality responding really is, which fraud signals dominate, and what a layered validation pipeline recovers.
Do Higher Rewards Buy Better Data? An Incentive Experiment
We fielded the same survey at four reward levels. Higher pay filled quotas dramatically faster - but quality moved less than folk wisdom predicts, in both directions.
Survey Length and Drop-Off: Where Respondents Quit
Completion data from 890,000 survey starts: drop-off is not linear. It spikes at the first question, stays flat to 8 minutes, then compounds ~2.3% per additional minute.
How Trust Scores Predict Future Response Quality
A participant's history predicts their next response better than any in-survey check: AUC 0.83 alone, 0.91 combined with per-survey signals, across 87,402 participants.
Detecting Straight-Lining: A Threshold Study
We tested straight-line detection thresholds from 0.70 to 0.95 across 320 Likert-heavy studies against a human reviewer panel. 0.85 wins - here is the full trade-off curve.
Attention Checks: Cost, Benefit, and Best Placement
A 6,000-participant randomized experiment: two attention checks catch 78% of low-effort responders at negligible cost. The fourth check catches almost nobody - and damages your data.
Mobile vs Desktop: Does Device Change Response Quality?
210,000 responses compared: mobile answers are faster and shorter, but once quality-scored, statistically indistinguishable from desktop. Do not down-weight mobile.