Honestly, anyone who isn't showing a 95% confidence interval is just being lazy. If you're not willing to run an experiment twenty or thirty times, you might as well not even look at the results and just publish it regardless of what it is. Why even research?

On the other hand, if you'd like to showcase actual insight and meaningful, surprising conclusions, it only makes sense to take the time to find the best dataset that supports it. Real researchers leave nothing to chance.