Hi, 

Are you talking about 'confusion' or 'confounding'? 

Shall answer your query after you clear my confusion! 

Best

Jeeva 

Some textbooks classify biases as selection bias,information bias,and confounding.

bias. Confounding bias is the same as confounding (not "confusing")

a confounder is an extraneous factor which is a determinant of the outcome and is assymetrically distributed between the  group exposed to the study factor and the reference group and which distorts the relationship of interest..

selection bias occurs when the property under study influences the selection of those into the study and information bias occurs when there is assymetry of information in the study and reference groups

Even though the origins of traditional biases like selection bias, recall bias, etc are different, usually confounding is also mentioned in the lists of "bias", because it distorts the estimate of the association between the exposure and the disease. Technically I would say it is "confounding"  and not confounding bias but both terms are used and definitely refer to the same "problem": where some unmeasured variable is associated with both the exposure and the outcome and distorts their association.

they are the same, but different books have different names. 

Dear

Bias creates an association that is not true,but confounding describes an association that is true,but potentialy misleading.

Have a look to this definition of "bias" in Nieuwenhuijsen MJ "Exposure assessment in occupational and environmental epidemiology." Oxford, Oxford University Press, 2003 - page  17.  

"Deviation of results or inferences  from the truth, or processes leading to such deviation: any trend in the design, collection, analysis, interpretation, publication, or review of data that can lead to conclusions that are sistematically different from the truth."

Bias can so create a figure that doesn't exist.

Confounding never creates ad image that doens't exist, but can obscure  a true image (like fog does), clouding figures vs. background and boundaries between really distinguished figures (e.g. when studying: "ever exposed", "occasionally exposed" and "never exposed"  workers; groups of persons living in areas respectively associated with "high" and "low" environmental pollution; a cohort of workers compared with the "general population" resident in the area where they work or worked) .

A particularly relevant confounder is a non-systematic misclassification of individuals whose profile should conduct to enrole them in a study group, while they are enlisted in another one.

The difference between groups rarely responds to the logic "white or black", in the practice we normally cope with scales of more ore less deep greys: if some "dark" individual is intermixed with the "pale" ones and vice versa, easily we'll find in front of a scenario of simply more dark or pale shadows, more difficult to comprehend than the real one.

Clearly, the effect of confounding  is more dangerous when the differences between groups are slight (i.e., in an epidemiological study, when the relative risk between a population under study and the reference lies around 2 and / or the lower limit of the confidence interval is near to 1).

I think the attached article could be useful to you.

Bias may arise during the implementation of the research, but "confounding" is in your mind. If you explore the confounding factor and clarify your mind, you can reach the truth easily.

As some of the colleagues indicated, classical epidemiology books and teaching classify bias (systematic error) as selection, information/classification and confounding. However a confounding effect results when the researcher ignores its role on the outcome. Confounding can be reduced at the stage of study design by matching or at the time of analysis. The available statistical methods take care of these issues.like multiple and logistic regression analysis.

I think good reading in the standard textbooks of epidemiology will further help colleagues to understand the  term "confounding" 

Confounding bias is the fraction of bias due to a given confounder. Confounding may be considered as the process of introducing bias by a confounder.