I know, that results from multiple linear regression, partial correlation analysis and ANOVA are tightly related to each other. The answers to this and this question indicate that the results may be "converted" to each other if appropriate information is given. However, for a concrete case, I am struggling to do so:
Suppose a multiple linear regression problem from an electrophysiological study where the latency of a measured nerve response is modeled by age, height and sex (dummy coded as 0/1).
$latency = \beta_0 + \beta_1 * age + \beta_2 * height + \beta_3 * sex$
From a research article, I obtain the partial correlation between age, height and sex and latence, the overall mean latency and its standard deviation. Additionally, I have the results from an ANOVA (SS, MS, df, F and p for independent each variable).
With these information, is it possible to obtain the multiple regression coefficients $\beta$? If yes, how?
