This is for you data nerds who might want to dig into the data a little deeper (we didn’t have a lot of time to play around with it). We’ve provided access to the full excel spreadsheet (which also includes data that was cut off from our analysis due to us having to get started earlier for today’s piece on BoingBoing), as well as the ongoing data published via google sheets.

We’re essentially hoping that with more rigorous analysis of the data, we might be able to find out whether there are statistically relevant differences in how different folks rate their candies For example, do people who see the dress in a certain way, more likely to have joy over certain candies, when compared to the other dress colour choice? Does knowing J.J. Abrams personally result in a certain portfoilio of candy choices? The types of analytical queries are kind of endless, and they don’t even have to relate specifically to candy.

Anyway, if any of you folks are game on, do let me know what you do with the data via twitter or email (db@mail.ubc.ca), and I’ll do my best to be post these other analyses up here at the SCQ – sort of creating a analytical hub if you will. Hopefully, we’ll be able to make this thing more and more robust each year.

Finally, an afterthought: something Ben and I noticed right after we sent the piece in. If you look at the Peterson Inflex and compare its position to last year, it’s a lot higher up the list. In a way, it sort of suggests that there was less joy overall this year. Might be curious to see how things look next year…


(LINKS TO – xlsx | published google sheet data)


Cohen B.R. and Ng. D. Only you can determine what tops the official Hallowe’en Candy Hierarchy. Boingboing, October 15th, 2014 – link

Cohen B.R. and Ng. D. The Candy Hierarchy 2015: your essential guide to tonight’s treats Boingboing, October 31st, 2015 – link


It’s always about the candy. The Candy Hierarchy is full up with this “joy induction” measurement, this thing that the co-principle investigators (PIs) Cohen and Ng go on about each year. From 2006 to 2013, the PIs conducted a longitudinal study, more or less guided by PI expertise and whim (or whimsical expertise) and possible corporation sponsorship. Research by others in the field sought to refute the findings, obviously unsuccessfully. Yet the PIs were so moved by the yearly outpouring of commentary that they opened up the study to additional data sources, namely people. People who the PIs surveyed. Or is it whom? Anyway, nobody cares – this is about sugar. The 2014 Candy Hierarchy was thus defined by data analysis of 43,767 votes obtained from 1286 individuals. Good for them. But not good enough for science. Because the 2015 Candy Hierarchy doubled down and reworked the whole thing with all kinds of more stuff. This hierarchy therefore presents the newly calculated 2015 rankings, based on a total of 518,605 data points obtained from 5459 individuals in a randomized fashion. It also provides the raw data from a secondary study that sought to understand the character of the survey takers, or rather how character affects joy induction. It’s all in there, just go check out the figures.


S. Tonev & B. Cohen. Stratigraphical Analysis of Friday-Sunday Identification in Relation to Sugar Consumption Preferences and Related Demographic Variables. The Science Creative Quarterly. November 2nd, 2015. (link)

J. Bryan. The joy of free text fields. Non-random sample of ~6K answers to “What is your favourite font?” November 5th, 2015. (link)