Class 8: Statistical interpretation and bias

Bias

Systematic errors or deviations from the "truth"

Type I error

Reject the null hypothesis when it's actually true

Type I error

Fail to reject the null hypothesis when it's actually true

Sampling error

Some differences between the sample and the population of interest

Sampling bias

Systemic error in sample selection

Confounder bias

Association is related to another variable other than then exposure or intervention

Hawthorne effect

Participants act differently when they know they're being watched

Confirmation bias

Only using results that support previous held beliefs

Recall bias

Humans are susceptible to incorrectly remembering

Attrition bias

As a study goes, some participants leave (loss to follow

up)

Publishing bias

Results that are statistically significant may be more likely to be published than non-significant results, leading to an overestimation of the true effect size.

Impact factor

Average number of citations to articles published in science and social science journals

Predatory journals

Poor peer review, spam researchers, low or nonexistent publishing standards, falsified editorial board, charge high fees for work

Selection bias

Research samples can sometimes under-re[resent certain people or groups

Restriction

“Your restrict your treatment group by only including subjects with the same values of potential confounding factors.”

Matching

Select a comparison group that matches with the treatment group.

Statistical control

Include the possible confounders as control variables in your regression models; in this way, you will control for the impact of the confounding variable.