Trust and honesty are important, but let's not confuse an interest in those with any specific type of research practice. Instead, let's address what seems like an elephant in the room, which is that eliminating 'research sins' is neither a necessary nor a sufficient way to produce engagement with worldly problems in useful ways. We can debate our own self-created and self-identified troubles as they relate to a particular image of science, but if this doesn't allow us to tackle worldly problems in useful ways, then what's the point? Our problem isn't being somehow more scientific--we've already taken that to an extreme--our problem is doing more useful work.
Along with many others, I am in favor of a very empirical science. In the name of such a science, it seems important to recognize that it offers insights that diffuse an interest in scientific truth, knowledge, or fact as kinds of abstracted 'things' of some kind. To be clear: there is consensus among scientists that the human being is a species of African great ape who develops within a culture to act, speak, and think in always-local ways. Terms such as 'truth' or 'knowledge' are words that this species uses, and words acquire their meanings in a way that is dynamically mapped onto social situations. Such dynamism in word usage is useful because, like other related animals, humans use language to organize collective action. As an animal, when we study the human at work, we find that it uses words such as truth, knowledge, or fact merely to coordinate its collective activity. Indeed, there really isn't anything spectacular about the language scientists use outside of how they coordinate their activity, including perceptions, beliefs, and thoughts, which are all active and learned social processes. Further, when we actually take an empirical look at the historical emergence of the sciences and their notions of truth, knowledge, or fact, we find that these grow out of a religious occupation with divining the essence or will or a deity in order to conform with it--the biblical laws very clearly shifted to become 'natural laws' in the 1600s. The evolution of the interest in a kind of transcendental divinity is still found in philosophy and the sciences when researchers look for something beyond their collective actions and discourse to abstractions such as 'truth' (or whatever)--you can see the worship here on anandtech. However, like all words, empirical science tells us that this is just another mundane utterance of an African great ape. There's really nothing spectacular about the words of scientists or anyone else as long as we remain empirical and scientific, rejecting a naive belief in the transcendental and quite religious character of any specific element of human discourse.
Further, regarding hypotheses, models, theories, and whatever else we're supposed to be chasing to be more scientific, it has long been recognized even by physicists that an infinite number of hypotheses, models, and theories can explain the same data equally well. This and related issues help to obliterate the possibility of arriving at a singular kind of truth or fact. This and related issues is discussed in various ways under terms such as 'meaning variance', the 'underdetermination of theory', the 'Quine-Duhem hypothesis', the 'problem of induction', and the list could go on. All such problems illustrate that the best and brightest among us have showed the impossibility of arriving at some abstracted singular thing called truth (or whatever). If we can get away from the transcendental addiction and become empirical scientists, we can recognize all forms of hypotheses, models, and theories as being merely types of human discourse, serving as analogical tools that an African great ape uses to reason and coordinate collective action. No form of this animal's discourse offers a path to a transcendental or otherwise divine location--no one gets to leave the cave, much less return to it. Unfortunately, as we have found in the organizational sciences, the practical aspects of the human's discursive activities are too often lost, forgotten, and misunderstood when it focuses on the analogical tools as gateways to something transcendental, such as a singular kind of truth (or whatever).
A profound empirical science is here if we want it, but in my view it would direct us away from our interest in 'science' itself as an abstraction, as a word that we would use to regulate our own activities in ways that take us away from the potential practical application of what we can collectively do and say. Is honesty important? Sure, but not because without it we risk "distorting reality" in any naive sense. Instead, we risk failing to collectively organize ourselves in useful ways, because telling the truth means playing by a set of local rules for collective discourse and action that require being able to trust each other. Instead of merely trying to enforce these rules, we should stop and consider whether or not the rules allow us to do good and important work. What is 'important work'? I'm not sure, but talking about it seems like the right conversation to have.
As a few examples, perhaps we can recognize that solving any research problem will do nothing to stop global warming, reduce social and economic inequalities, eliminate corruption in business and government, increase democratic participation, and the like. By focusing on being more 'scientific', we risk forgetting that we're members of a human species. In order to achieve our great potentials, earn our large salaries, and grapple with useful means and ends, it seems time to start speaking and acting differently. A staunchly empirical science, as a collective human endeavor, should have us do no less.