Finding out whether a decision made has been risk-taking or risk-averse

Question

I want to do an empirical study on risk management by translation trainees for my MA thesis. Translation is an act full of decision-makings for rendering risky parts of the text. I want to interpret each decision the subjects make in their own context, and I believe the only tool I have to assess whether the decisions made are risk-averse or risk-taking is to interpret justifications they provide for the decisions they make (both when thinking out loud and in the interview I will have with them after the task). But I need a framework for the basis of such interpretations; I guess something that specifies A, B, and C types of arguments in their justifications show that their decision has been risk-taking, and X, Y, and Z arguments show that their decision has been risk-averse. Something that can be applied to each decision made, not an overall approach to the decision-maker's characteristics. Do you suggest any theoretical framework for such work?

P.S.: Risk in translation? Well, some believe risk is "The probability of a desired or an undesired outcome as a consequence of a (translational) action". Pym also argues that each text consists of high-risk, medium-risk, and low-risk items, e.g. the title in a letter is a high-risk item (very important in communication) but conjunctions are low-risk items. Pym also raises this idea that translators deal with 3 kinds of risks: credibility risk (keeping/losing the reputation as the translator), uncertainty risk (about HOW to translate items), and communicative risk (which is related to the text itself). On the other hand, some scholars believe that each element of a text can have political, economic, etc. type of risk within itself, but they do not raise this idea in a structured way.The latter group believes that for example if you are working for a Pan-X newspaper and translate the item Y in a way which is opposing the ideologies of the newspaper, you are taking a risk, because it can make you get fired. [Which sense you find reasonable, by the way?]

Answer 1

I would recommend establishing (quantifying) risk tolerance first. Risks accepted near or above the risk tolerance are risk taking; risks rejected near or below the risk tolerance are risk averse.

So the trick is to find a way to calibrate your measurement of translation risk. Sounds like you have a beginning - differentiating between titles and articles and adverbs. I don't know the field very well, but I suspect you could begin by generating a set of possibilities for key words in translation, and then measuring the improbability of the choice. I suspect that you could generate markov chains of word sequences in both languages. I remember that there is some pre-existing research that enables you to measure the distance of a word pair from a baseline, but I haven't looked at the topic since the last century; I'm confident you can find something relevant to your field and use that to establish a baseline risk tolerance - some translators will choose conventional translations, while others will select non-conventional terms in an attempt to convey subtle nuance. The non-conventional choices represent higher risk. My hypothesis is that each translator will have a pattern - a range of risk in translation. That also permits you to identify translation choices that significantly deviate from the comfort zone; those can be classified as risk averse (closer to convention than to the translator's baseline), or risk seeking (further from convention than the translator's baseline. Those become the center of interviews.

I'm well aware of my ignorance, and I hope that I haven't offended you by over-simplifying your research. I'm trying to articulate something that is partly lost to age and memory and partly something that exceeds my competence. Good luck.

Aside 1: there is ample scientific research to indicate that the NIST 800-37 risk management standard ("high", "medium" "low") has negative value - differing calibrations of those terms actually interfere with communications and preclude risk management. Avoid this if you can. If you must use qualitative risk management, calibrate it carefully so that independent observers reach the same conclusion.

Aside 2: you may wish to consult the society of information risk analysis. You may also wish to research Tony Cox's writings on risk. They aren't referring to project risk, but to information security risk, but they have a more academic, more formal approach to risk management.