Statistics studies
From wikinotes
Conducting a Study
Steps in conducting a study
- State a hypothesis
- Identify individuals of interest
- State variables to measure
- Choose entire-population VS sample
- Address ethical concerns before collecting data
- Collect Data
- Use descriptive or inferential stats to answer hypothesis
- Note concerns about data collection/analysis
Study Types
Experimental: Change in behaviour is prescribed, and measure the outcomes Observational: Don't change behaviour, just observe
Considerations
- Studies must be rigorous enough that they can be replicated
- Bias
- Non-Response/Voluntary Response (who is not responding? is this a fair distribution?)
- Truthfulness (are people lying on purpose? Are questions too personal?)
- Recall-Bias (events that occurred a long time ago, mis-remembering events)
- Hidden Bias (does the wording of your questions influence how people reply? if grading 1-5 is that fine-grad enough?)
- Interviewer Influence: (should you choose an interviewer that is more similar to your sample?)
- Placebo: including a placebo group helps account for the placebo-effect
- Blinding: ensure participants, and those conducting study are unaware of control group they are in (double blind)