Last year, consumers in the United States expected to spend an all-time high of 12.2 billion U.S. dollars for the Halloween season. In 2024, however, U.S. shoppers were expected to spend about half a billion less. In fact, average spending per person was expected to drop by nearly five dollars when compared to the previous year. Shopping on All Hallows’ Eve From terrifying jack-o’-lanterns and skeleton props to chilling face makeup and treats, there are various items commonly purchased by consumers during the Halloween season. In 2024, Americans were expected to spend a total of nearly four billion U.S. dollars on costumes and outfits alone. Another three and a half billion U.S. dollars was projected to be spent on candy. Trick-or-treat! In 2024, over seven in 10 surveyed individuals in the United States said they would be celebrating Halloween, whether it is in the form of dressing up for a themed party or going out at night to trick-or-treat on or around October 31. When trick-or-treating at night, one never knows what treats one might get (e.g., licorice, chocolate, or maybe even money). But the households that do not offer anything to trick-or-treating souls can sometimes expect mischief to occur on their property.

Context

This is a Halloween survey from data collected via Facebook, LinkedIn, Reddit, and email. Respondents were offered Amazon gift cards as random prizes. The questions include what costumes people wore, when people started/stopped trick-or-treating, a few other Halloween-related questions, and demographics.

Content

The responses are all (ideally, though not verified) from people who have spent most of their lives in the United States/its territories.

There are responses for individuals who did and didn't celebrate Halloween, as well as those who did but never went trick-or-treating. The sample may possibly be biased in that regard, although I hope that the gift card prizes reduce that effect.

The first main section is with costumes people wore. There is a checkbox grid question with 4 groups for each of 35 different costumes/costume categories. The options correspond to "Age 9 and under", "Ages 10-17", "As an adult", and "One of these, but I don't remember".

The second section asks for trick-or-treating ages: when they started, and when they stopped. If respondents could not remember, they could put down an interval, such as 3-5 for starting trick-or-treating, and 12-13 for end.

Each response was also asked if they attended haunted houses at various ages, as well as a few words to describe Halloween. Favorite and least-favorite candies were also free-response questions.

The last section asks for demographics. The regions that are described correspond to the ones here: https://www.ilibrarian.net/flagmaps/us_states_map_regions_lg.jpg

The order of responses has been randomized from the original data sources.

Inspiration

Although the sample size isn't particularly large, I am interested in seeing what others may do with the data. I have already analyzed the data, with the code on GitHub here: https://github.com/mcandocia/halloween_survey

I have already analyzed some of the data by

• using survival analysis to determine when kids trick or treat

• using raking to estimate which costumes have been the most popular

• some wordclouds

However, these are only a few ways of analyzing a few things in the data. I'm interested in seeing what else others may come up with.