Question

Which of the following is NOT a type of bias?

-Randomization bias

-Misclassification bias

-Confounding

- information bias

- selection bias

Answer 1

Bias is like being bias to anything, either it can occur with sampling or with the statistic. Any error which occurs while building/ studying an association between exposure ( risk factor) and disease.

Randomization bias --- This is NOT a type of bias. Randomized control trial is a term, but it is not a type of bias.

Misclassification bias --- When study participants are misclassified with regard to disease or exposure is incorrect.

It is of 2 types --

1. Differential/ Non-random --- In this case, misclassification occurs non-randomly between the 2 sets of groups taken for eg. case-control and cohort.

2. Non-differential/ Random --- In this case, misclassification occurs randomly between the 2 sets of groups taken for eg. case-control and cohort.

Confounding --- Confounding bias occurs when a result misleads us to think that 1 variable causes the other but what is causing really is the third variable that causes our dependant variable and this third variable also happens to be related to our independent variable. So, the secondary factors sometimes lead to disease and when that is not being considered is known as confounding bias.

Information bias -- Study bias because of wrong information provided or gathered, as the patient is giving the wrong information or interviewer asking biased questions. It is of 2 types ---

1. Memory/recall bias -- When the patient is not able to recall the question asked and giving the wrong answer because of inability to recall the exact information or he is not able to remember the incident, this creates a wrong study basis and leads to study error and thus causing information bias.

2. Interviewer's bias -- When the interviewer is being biased while asking the questions so the result made leads to study error causing information bias.

Selection bias -- It occurs while selecting the population. While doing the study and selecting the participants choosing the wrong sample as the sample group not representing the entire target population.

It is too hard to study the entire population, so to ease this process a small sample group is created, and when the participant pool or the subsequent data is not representative of the target population this error is called selection bias. So, in order to prevent this bias selection of sample size should be done randomly.

Out of the above 5, 4 are types of bias. So, the answer is option A, Randomization bias is NOT a type of bias.