I would like to highlight a point on page 344 regarding the superiority of confidence intervals over tests and P values. Confidence intervals enable us to redirect our focus away from the null hypothesis towards the entire spectrum of effect sizes consistent with the data. This shift in focus is advocated by numerous authors and an increasing number of journals.
I find this point particularly helpful because, during statistical analysis, we often encounter situations where the p-value is significant but the confidence interval is not significant or encompasses 1. In such cases, it raises the question of what we should believe. This insight provides a valuable answer for me.