Hello Alka,

Your result means that, given your data set, you can not conclude that observed differences are a function of anything other than chance (if your threshold for statistical significance is .01 for the computed p-value). Therefore, you should behave as if the treatments are equal in effect.

Why might this occur?

1. There are truly no differences in the parameter of interest (e.g., mean) among the populations represented by your treatment groups.

2. There are differences, but your sample size was insufficient to detect them (because of low statistical power or just an "unlucky" sample data set).

Good luck with your work.

SIMPLE ANSWER. 1% significance means the error must be 1% or less and the remaining 99% is considered normal events. If you result is not significant at 1% it means that 99% of the event you studied is considered normal and your dinding is 99% or less therefore is it not significant....just normal occurrence.