Free Trial

Safari Books Online is a digital library providing on-demand subscription access to thousands of learning resources.

Share this Page URL

11.2.1 R Function reg2ci > 11.2.1 R Function reg2ci - Pg. 552

552 Introduction to Robust Estimation and Hypothesis Testing computes a 0.95 confidence interval for the difference between regression slope parameters corresponding to two independent groups. The first two arguments contain the data for the first group, and the data for group 2 are contained in x2 and y2. The optional argument regfun indicates the regression method to be used. If not specified, tsreg, the Theil­Sen estimator, is ^ used. Setting regfun=bmreg results in using m , M regression with Schweppe weights. The default number of bootstrap samples, nboot, is B = 599, and alpha, which is , defaults to 0.05 if unspecified. When the argument plotit equals T (for true), the function also creates a scatterplot that includes the regression lines for both groups. I Example A controversial issue is whether teachers' expectancies in uence intellectual functioning. A generic title for studies that address this issue is Pygmalion in the classroom. Rosenthal and Jacobson (1968) argue that teachers' expectancies in uence intellectual functioning, and others argue that it does not. A brief summary of some of the counterarguments can be found in Snow (1995). Snow illustrates his concerns with data collected by Rosenthal where children in grades 1 and 2 were used. Here, other issues are examined using robust regression methods.