Response styles are one of the major sources of common method bias. They refer to the tendency to select specific categories of rating scale. The most common types of response styles include acquiescence response style, disacquiescence response style, extreme response style and midpoint response style. They can cause subjects’ response to rating scale biased, and thus, they may lead to systematic measurement errors in test scores and influence the results of the analyses of test reliability and validity and the analysis of the covariation between constructs. It is hard to measure response styles directly with a psychological scale, and thus researchers usually use the counting procedure or the modeling procedure to measure it. The counting procedure mainly includes the traditional counting procedure, the method of representative indicators for response styles, and the method of counting double agreements on reversed items. The modeling procedure mainly includes the method of specifying acquiescence response style in confirmatory factor analysis, confirmatory latent class analysis, the mixed partial credit model, the item response tree model and the multidimensional nominal response model. When selecting appropriate methods to measure response styles, researchers need to consider the following questions. Firstly, what types of response styles they want to measure? Secondly, are response styles regarded as continuous or categorical variables? Thirdly, is the content-related trait regarded as a continuous or categorical variable? Fourthly, can researchers obtain a group of heterogeneous items? Fifthly, are there some pairs of items that examine similar content but are worded in opposite directions? Lastly, is the number of positively worded items equal to the number of negatively worded items? As for controlling the effects of response styles on data analysis, different procedures are appropriate in controlling different effects of response styles, and in some cases, it requires researchers to combine the measuring procedure and the regression residual method or the partial correlation method to control the effects of response styles. Firstly, to control the bias that response styles cause in test scores, researchers can apply the modeling procedure or the combination of the counting procedure and the regression residual method. Secondly, to control the effects of response styles on the analyses of test reliability and validity, researchers can apply the combination of the modeling procedure and the regression residual method or the combination of the counting procedure and the regression residual method. With respect to the control of the effects of response styles on the measurement invariance test, researchers can apply the procedure of specifying acquiescence response style in confirmatory factor analysis, confirmatory latent class analysis or the multidimensional nominal response model. Thirdly, to control the effect of response styles on the analysis of the covariation between constructs, researchers can apply the modeling procedure, the combination of the counting procedure and the regression residual method or the combination of the counting procedure and the partial correlation method. Future research needs to extend the existing methods for measuring and controlling response styles, examine the validity of them, and systematically investigate the effects of response styles on data analysis.