Negatively skewed rank orders

We tend to think of these choices as all being resonable and evenly distributed between good and bad choices, with possibly a couple of desirable responses, maybe some neutral ones, and a couple of less desirable responses.

The response sets below are designed for drag & drop ordering. Drag the most desirable response into the top slot. Order the responses by dragging them into your preferred order, top to bottom. 

Remember that within the SJT rank-order scoring mechanism, it is not just your first choice that counts. The order of your last 2 choices counts for just as much. 

But what about situations, like you might find in ethical dilemmas, where none of the choices are good ones, but circumstances force you to choose between the devil and the deep blue sea? 

Also we should consider weighting in that sometimes the choices are so nearly the same that the spread amongst your reference panel will be really flat, or some where the differences are poles apart. The former is not generally a good question, if it is largely a coin flip as to which you would rate over another; the latter is also not useful if it is really obvious as to what order things should be ranked. Finding the middle ground in terms of difficulty or discrimination is what we are generally looking for. Item Response Theory and Rasch modeling is somewhat helpful in this but there are caveats to these. 

Thank you