1. Asking for **reported** **data** instead of actually measuring it. 2. Asking **loaded questions** or asking questions in a **suggestive order**. 3. Asking for **voluntary responses** when that could introduce response bias (e.g., political polls). Use voluntary responses only if you _want_ to introduce the self-selection bias, to gather feedback (from your customers, listeners/watchers of a radio/TV show, your team). 4. Not checking if the **source** of the data **or** the **sponsor** of the study **have a** potential **conflict of interest** (Ask yourself if there is pressure on the source/sponsor to obtain specific outcomes.) 5. **Sample bias** based on time, space, or (survey) method. For example, surveying only at 13:00-14:00, in a location that only specific people tend to visit, or only calling registered, public phone numbers. Source: Elementary Statistics by Allan Bluman