The report discusses what fraction of employees are female, black, or Hispanic, but it does not mention the demographics of the hiring pool. If, for example, the majority of Google hires have undergraduate degrees in computer science, you need to know what fraction of CS majors are female, black, or Hispanic. Having a higher fraction female programmers at Google than are in the pool is possible only if the average female programmer is better or if Google discriminates on the basis of sex.
Comments
The report discusses what fraction of employees are female, black, or Hispanic, but it does not mention the demographics of the hiring pool. If, for example, the majority of Google hires have undergraduate degrees in computer science, you need to know what fraction of CS majors are female, black, or Hispanic. Having a higher fraction female programmers at Google than are in the pool is possible only if the average female programmer is better or if Google discriminates on the basis of sex.
...or a hundred other reasons, like their hiring process is problematic or there are other confounding variables.
But yes, you do need to know the prior statistics to understand the result.