For sure, blinding is an important tool for reducing bias and essential for optimum experimental validity. I'm more familiar with the medical field but the fun

For sure, blinding is an important tool for reducing bias and essential for optimum experimental validity. I'm more familiar with the medical field but the fundamental issues are the same. In recent years there's been growing concern about the limitations of blinding. In pharmaceutical research, the random controlled trial (RCT) remains the "gold standard", but over time, increasing strength of placebo response has become a big problem. IOW as separation of conditions is less affected by blinding, it ceases to be a useful experimental tool. Of course, this may not apply so directly to observations of insect behavior but it does illustrate that "blindness" to experimental conditions is relative, not absolute. The publication process could itself inject biases into experimental design and conduct. Having to pass through peer review potentially distorts things. If the experimenter believes reviewers favor particular methods or outcomes, it could influence the experimental set up or even whether the experiment is carried out. I've long thought one way to improve publication quality would be for editors to forbid authors drawing broad conclusions vs. just putting results into context. After all, "first findings" are prized, but such results are preliminary or unconfirmed. It will take many studies for genuine patterns to emerge. The conclusions have to come later as the community of science determines what findings mean. Avoiding premature speculation about conclusions will reduce bias because "negative" studies would then just be more data, part of the pattern to be puzzled out, and more likely to be published vs. relegated to the "file drawer". The quality of the study becomes more important than having to ascend to the top of the heap, to show my conclusions grander than the others. Eliminating that contest will improve the discourse immeasurably.