Many statistical analyses generate both P values and confidence intervals. Many scientists report the P value and ignore the confidence interval.

I think this is a mistake.

Interpreting P values is tricky. Interpreting confidence intervals, in contrast, is quite simple. You collect some data, do some calculations to quantify a difference (or ratio, or best-fit value...), and report that value along with a confidence interval to show how precise that value is.

The underlying theory is identical for confidence intervals and P values. So if both are interpreted correctly, the conclusions are identical. But that is a big 'if'', and I agree with the following quote (JM Hoenig and DM Heisey, The American Statistician, 55: 1-6, 2001):

"... imperfectly understood confidence intervals are more useful and less dangerous than incorrectly understood P values and hypothesis tests."