17
votes
Is it (always) better to build a model prior to viewing the data?
The problem is that of potential bias in p-values and confidence intervals as well as overfitting or poor predictive performance without making adjustments for the process of choosing the model. ...
- 811
16
votes
Accepted
Is it (always) better to build a model prior to viewing the data?
If you build a model before seeing the data you are more likely to get results (effect sizes, p-values, etc.) that correctly relate the variables in the model to the process you are studying. That is, ...
- 5,349
12
votes
Is it (always) better to build a model prior to viewing the data?
There is a division, or gradation, between exploratory and confirmatory statistics. Both have value. Mixing the two is dangerous.
At one end, we have a lot of data mining. Some of this comes very ...
- 128k
2
votes
Accepted
Would withholding marks until students respond to a survey bias the responses?
Satisficing is at least one problem to consider. For instance, people may simply rush through the survey as an answering strategy, in particular if the survey is long.
Non-response bias may be ...
- 5,848
Only top scored, non community-wiki answers of a minimum length are eligible