There have been many people who have looked at John's work, including well-recognized experts like Scott Burns. Never have I heard a single criticism from any reasonably informed person.

Flat-out lie. There have been many informed people who criticized JWR's work - you just re-write history.

I'm "reasonably informed" regarding statistics (although I claim no expertise in economics) and use statistics almost every day. On the old NFB, I asked JWR some simple questions, such as...

1. How did he justify deleting data from his analyses? Competent, honest analysts always have objective criteria for deleting data and publically describe them. That's one of the most fundamental ways to keep statistical analyses honest.

2. Is it valid to "eyeball fit" confidence intervals? Confidence intervals are well-defined statistical parameters that are

calculated. His guesses are, quite simply, meaningless.

3. The r2 values of his correlations were so low (0.2 - 0.4) that I questioned the useful of predictions from his equations.

4. When I calculated the r2 values, I got even lower values than he reported - even though I used the same data set that JWR used. I asked him about this.

JWR's response to this and other questions, of course, was to simply run away and refuse to respond.

Now, hocus, tell us. In this stock market predictor of yours, how have you and JWR validated the fundamental assumptions of linear regression that determine whether

any of your predictions have value.

1. Linearity of the relationship between dependent and independent variables.

2. Normality of the distribution.

3. Homoscedasticity of errors

4. Absence of serial correlation.

After that, you could address details such as r2 values, "eyeball" estimates of correlation coefficients, etc.

Naw...never mind. You're incapable of responding honestly even if you or JWR knew the answers.