Significance of bootstrap results

[soumyaray]

DO NOT MERGE

Playing with some methods that relate to bootstrapped testing. All implemented methods must be considered tentative and noted in documentation as 'experimental'.

Tentative solutions:

• Test of difference between path coefficients within same model (abs_diff, tvalues, pvalues)
  • quesions remaining:
    • which t distribution to check pvalues from (alt/null)?

Upcoming:

• p-values of path coefficients
  • questions remaining
    • use quantile of tvalues?
    • which t distribution to check pvalues from (alt/null)?