The Burrito Blog

« The Globe's Blind Burrito Taste Test | Chipotle Giveaway: Harvard Sq. 3/12/08 »

Burrito Half Life And Causes Of Error

There's been a whirlwind of hate mail peer review regarding the recent burrito taste test organized by The Boston Globe.

"I am shocked and also offended, perhaps next time you should include a real burrito expert, as in myself."

"I am sick and tired of you blasting Boca Grande. You obviously do not have very discerning taste buds."

"No Felipe's? Fail."

Despite hostile tones, I share a passion with you, fellow burrito analysts. I hear your unrest. And I understand. No taste test is perfect, and we must consider the flaws of the experimental design.

#1: Transportation, Preservation, and Heat Loss

Some burritos were purchased up to sixty minutes in advance to account for transport to the testing site. Preservation techniques were limited to paper bags for the Friday rush hour journey to Faneuil Hall. Some burritos were cold.

#2: Evaporation Catalysts
Moisture is critical to tortilla texture. Some burritos were incised widthwise to prepare for photography, lasting approximately four minutes. Prolonged exposure to air accelerated the rate of evaporation prior to tasting. Tortillas were drier than usual.

#3: Burrito Half Life and Quality Decay
Anecdotal estimates place optimal consumption within seven minutes of construction. To supplement scarce scientific documentation on burrito decay, our analyst team composed the following estimate:

Because of varying purchase times to account for travel, freshness was unequal.

#4: Configuration as an Uncontrolled Variable
Efforts were made to ensure consistent burrito specifications as follows: steak, pinto beans, rice, cheese, and salsa. However, the Boca Grande specimen had no salsa, and varying hot sauce application techniques made controlling for spice difficult. It is impossible to configure burritos identically.

#5: Untested Vendors
The taste test was limited to five burritos; numerous local taquerias were stricken from participation.

#6: Natural Variance and Small Sample Size
Hand made burritos are subject to unavoidable inconsistency. When comparing the means of two samples with high variance, large sample sizes are required to warrant conclusions. Each tester sampled only one burrito from each establishment.

#7: Absent Rating Criteria
Burrito Blog's rating system incorporates cleanliness, presentation, service, and how I feel four hours later. The nature of the blind taste test at a neutral location invalidated the first three of these measurements. Mixed consumption and immediate judging eliminated the most important rating criteria: the aftermath.

Despite extensive causes of error, we applaud The Boston Globe in their scientific quest. Detracting elements were entirely beyond their control.

As with all experiments, repetition is required. So, dear readers, go out and eat for yourself.

This entry is filed in {Press}.

Sunday, March 09, 2008 by Jonah.