How do we create our questions?
There is a whole team of incredible people behind people.io whose work often goes unnoticed by our community. We therefore thought we’d ask them to explain how they’ve personally applied their awesomeness to make your people.io experience as magical as possible.
Today’s blog post comes to you from our resident neuropsychologist and jamón lover, Sonia, who goes into detail about the thought processes behind our questions:
What makes you, you?
We always say it: we want to give people ownership of their data, but what exactly is your data? To give you a few examples: your data can be what you love, what you hate, how you behave, how you feel and how you think. What makes this even more interesting is that your data is never static: what excited you yesterday may not excite you today, and the brands that shaped your 2017 may not make the 2018 cut. Your situation changes constantly, and the data that defines you is just as dynamic.
When we create questions for the people.io community, we try to address each of these challenges, and more. If you’re curious about the method behind our approach — read on.
What questions do we ask? Why do we ask these questions?
Our questions are designed to slot into a hierarchy of categories. At the top of this hierarchy, we have broad categories such as demographics and interests. These then branch out to become more and more specific.
The way we ask questions around these categories follows a psychological approach known as the ABC model, which stands for Affective, Behavioural & Cognitive. In simpler terms, this takes into account how a person feels, behaves and thinks when presented with certain topics. This means that every time you swipe agree or disagree, you’re being positively or negatively associated with different topics related to that question e.g. liking or disliking pop music.
From these associations, we create a network similar to that found between neurons in the brain. All neurons in your brain are interconnected, whether directly or indirectly, but the strength of those connections varies. As an example, in our model, the ‘neuron’ that reveals your love of hip-hop will be strongly connected to your Cardi B ‘neuron’. Of course, this connection will need further evidence and data to make it stronger and keep it fresh, just as your brain takes in different stimuli to strengthen the connections between your neurons. In short, the more questions we ask you, and the more frequently you respond to them, the more well-defined and accurate a picture we can maintain of your network of interests.
But just how do we detect how strongly associated you are with each of our categories? For that, we need to ask questions based on mental models and cognitive concepts such as attention, memory and decision making. That’s a lot of neuropsychology for one sentence, so here are some examples to help explain:
Attention — How aware are you of global warming?
When designing our questions, the first principle we consider is attention, that is, how aware you are of a certain notion, person or situation. If your levels of attention are low, this means that you are simply aware of something’s existence. High levels of attention indicate affinity or loyalty, and show us that you are truly engaged with that thing.
Questions we may have asked to determine your attention include: “I know who Pharrell Williams is” or “I’ve played Candy Crush”
Affection, Reward System, and Decision Making — How excited are you about chocolate?
Once we’ve tried something, our brain consciously or subconsciously labels it with any range of emotions from good to bad. These labels tell us a lot about how you as a person are incentivised, (also known as your reward system). This is because, when you like something, your levels of reward hormones, dopamine or serotonin, increase, and make you want more of that thing. If we know that you keep coming back to that same pleasure-inducing stimulus, we can be more confident about your association with it.
Questions we may have asked you that take these principles into account include: “I can’t stop listening to Dua Lipa” or “I hate chocolate ice cream”
Memory — Have you been grumpy in the last week?
Our memory is limited. Traditionally, it’s been thought that an average person can only hold between five and nine objects in their working memory, (a theory known as the magical number seven, plus or minus two), but current research suggests even this is optimistic.
Our ability to retrieve our memories is dependent on many factors, including how long we’ve been exposed to something, how much attention we’ve paid to that thing, and how that thing made us feel. Within this, there are also certain biases that cause us to process information inaccurately.
This is often difficult to consider when creating questions, but we do take it into account. For example, you’ll find yourself being asked about a particular theme in a few different ways, or being asked the same question again weeks or even months later. This validates the strength of the relationships within your network of interests.
Questions we may have asked you that take memory into account include: “I regularly socialise with groups of friends”, “I go out with my friends at least twice a month”, “Last weekend I went out with friends”
Updated circumstances and self-perception — Is Harry Kane still your idol?
As humans, our circumstances often change, and so too do our needs, interests, behaviour and personality. Throughout life, the brain deals with these changes by re-organising itself and forming new neural connections, a concept known as neuroplasticity. This is another reason as to why you’ll see some questions being asked again after a certain period of time — you’re dynamic, and so is your data.
Questions we may have asked you that take these principles into account include: “I’m trying to cut down on meat”
What does this mean for you?
The more questions you answer with people.io, the more personalised your experience becomes as we understand more about your network of interests. Over time, this means that you will only be asked questions and matched with brands that are relevant to you.
As with everything we do, you are at the heart of this process: We never sell or share this, or any of your data with brands. Your data is valuable, and that’s why we reward you for it.