Stopped reading after I saw the chart drew lines to extrapolate from extremes relatively close to the mean hours of sleep, and accounted not for such things as degrees of standard deviation, rather (mis)leading to the conclusion that these happiness results were dependent entirely on the one measured variable and not that other factors would contribute to outliers. Your anecdotal evidence, and analysis, are flawed, though I lack the terminology to adequately improve them. Will someone please translate my hunch to appropriate statistical tools? I would like to replicate this data collection in my own life and learn something of value from it. My comment may find itself below a vast wall of anecdotal text, however, I’m hopeful, HN.