I'm always fascinated by questions. What questions to ask, how to frame questions, what kinds are best and so on.
Lately, I’ve been thinking about questions that trigger the response– "It depends"– as I hear this a lot and sometimes you may catch me saying it too.
A few years ago, my default reaction was " ‘It depends’ is a cop out answer ". But now I have shifted my thinking to blame the question format rather than the answer.
In a nutshell:
"It depends" means the question is context dependent and requires more precision. This means the more precise the question, the better the answer, which can eliminate "It depends".
Given that I am in the design field, I will use some design examples to illustrate the relationship between precision and "it depends".
A person new to the UX Design world could ask the following question:
"Which UX design tool should a newbie use?".
If the first thing that came to your mind was "It depends", then you have travelled well along this trodden path. In this example, we can infer that the answer depends on what the newcomer wants to achieve, depends on their background, depends on the company they work for and so on.
But if we increase the precision of the question to..
"What is the most popular prototyping design tool for iOS in the past year?", we increase the level of precision which does not give much room for "It depends". [1]
Let us compare the answers of the question using Bing’s Copilot on GPT-4.
Q: "Which UX design tool should a newbie use?"
GPT-4: “There is no definitive answer to which UX design tool is the best for beginners, as different tools have different strengths and weaknesses, and the choice may depend on personal preference, budget, and project requirements. However, based on the web search results, some of the most popular…”
Full answer is at the bottom of this essay. Emphasis on “may depend” is mine as I wanted to highlight it. [2]
Q: "What is the most popular prototyping design tool for iOS in the past year?"
GPT-4: “According to the web search results, the most popular prototyping tool for iOS in 2020 was Sketch²⁴. Sketch is a vector-based design software that offers a simple and lightweight interface, numerous plugins and integrations, and a wide range of features for creating user interfaces and prototypes¹².”
Full answer is at the bottom of this essay. Emphasis on Sketch was actually done by GPT-4. [3]
As shown, even the GenAI is able to give more direct answers when the precision is increased.
Given that we have established how precision works, the item to address is “What are the key ingredients that give precision to a question?”
Having thought about it, the main ones I believe are key are:
One of One
One of One refers to using words that point to the uniqueness of the context. In the example above “newbie” and “iOS” are unique items. Broadly speaking, there are no variations of these words, so it is hard to confuse them for something else, hence they are one of one.
Superlatives
Using the highest order of an adjective or adverb helps to bring focus on the type of answer required for the question. In the example above, “most popular” brings focus on the type of output required.
Discrete Numbers
Discrete numbers are similar superlatives as they often entail the highest order but not always. The newcomer's question could have been phrased as the top 5 or number 1 design software. Discrete numbers are important because there are many words to describe items greater than 1 (e.g. various), which can reduce the precision.
Boxed in boundaries
This is usually time or space bound. It could have a start or finish or could be environmentally fenced in. Putting boundaries in your question greatly reduces ambiguity. In the example above, “in the past year” is an example of a time boundary.
Important to note that these ingredients do not all have to be present in the question but it helps to have a few in there.
How is this useful in the real world?
If someone asks you an ambiguous question that would trigger the answer “it depends”, you can always reframe the question using the ingredients above.
For example:
