| answer to your question then some editorializing on this thread: answer: yes there are better ways to independently measure mental health than one self-report variable. Generally in the social sciences you can measure things in a number of broad ways. Generally we refer to a 'construct' - a theoretical description of some phenomenon or set of concepts that is difficult or impossible to measure directly. Basically, it is a definition and the hard part is that you can't directly touch the thing so sometimes these 'variables' have definitions that get argued about in the margins - c.f., motivation, how do you define categorize and measure motivation? Effectively, we are dealing with ways of evidencing things that are inferred. There are a couple paradigms of measurement (more than the two I'll mention): Observation - using some protocol or clear and as objective as possible tool for having a (typically) trained external individual. This can be either an overarching conclusion or observation of specific actions or symptoms: Self report - relying on the person reporting things. This can be the constructs themselves (not great) or it can be more concrete markers of things (better but still used with caution. The issue here is sometimes lying, sometimes social norming, sometimes self-awareness - all represent sources of error. Asking self report questions directly isn't inherently bad, but for something like mental health which is a latent construct, a single direct self report question is not going to get you accepted into any serious journal. You might ask 10 or 15 indirect, symptoms based questions, that is called a scale. For example you would ask: "how often over the last two weeks have you felt little interest or pleasure in doing things?" (from the PHQ-9, which implements the DSM-IV clinical criteria from depression). personal opinion: The question that the grandparent mentions is "In general, how has instagram affected the way you feel about your mental health?" Asking teenagers that is just...not great research praxis. The reason asking that is not great rather than bad is because it was part of a battery of data that they collected, all of which has to be interpreted together. For a lot of reasons, likert scale questions like this are bad, particular bad with children (who will tell you what they think you want to here), and increasingly bad in a society where 'please rate my customer service a 10 because a 9 is failing' is the norm. In fact, most respected likert-like scales still see a bias in all answers above the norm - a distributional skew towards the 'good' option whether its agree or better or whatever the scale is. The results of this question alone are not nearly as positive as they are implied to be. The next slides begin the absolute death for the claim about slide 24 ... Slide 25 - explains that the positives come from memes and comedians. That drop to 'conversations you have' and 'comments you receive' absolutely wipes out the idea that instagram is a positive. The sorting of this group of categories makes me honestly just sad. Slide 26 - "While the overall effects of instagram are positive [ed: that is not how I interpret this data], the effects are determined by the moment." I would encourage you to read the quotes. Slide 27 - "Teens blame instagram for increase in the rates of anxiety and depression among teens" Take a look at the text of this one as well...kids aren't stupid Slide 28 - "Teens who struggle with mental health say instagram makes it worse" What collectively this means is - it amplifies. For developing children who already have unstable hormonal and neurological responses, that is like society giving an arsonist a lighter and me claiming its not my fault because I only gave him accelerant. The teens understand this and the data backs them up. If your consultants bother to categorize the harm your company is doing pulling out one question and oversimplifying the results isn't an effective counter argument. Not only is there a better way than a single self-report variable - this variable was just plucked completely out of context and a claim made about it that is unsupported. |