Comment by shakna

The p-value is useless, where the t-value does not hold substance. One depends upon the other. If there's too much of a degree of freedom, it doesn't matter if the p-value looks accurate. The data is probably no longer normally distributed, requiring non-parametric testing.

You've leapt to me being a researcher acting in bad faith, accusing me for a whole industry. As to defunding an entire field of research, it sounds like you'd like statistics or mathematics defunded? I'm afraid they will persist regardless. Too many industries depend upon them.