Methodology: “Shmutz” Survey of Subway Car Cleanliness

Background

This report is intended as a follow-up to the April 1998, February 1999, January 2000 and February 2001 NYPIRG Straphangers Campaign surveys, Subway Shmutz: Cleanliness in New York City Subway Cars. The methodology followed in this survey is largely identical to that used in the 2000 and 2001 reports. One adjustment to the methodology of the 1998 and 1999 surveys is noted below.

Preparation

In July 1997, August 1999 and again in September 2002, our project directors accompanied members of the New York City Transit Passenger Environment Survey (PES) staff on an in-field mock rating of some 15 cars on two subway lines. This session included clarification of New York City Transit’s own rating system (described below). We also received the actual survey form used by PES staff. The PES looks at several measurements of the subway car environment, including car and seat cleanliness. Our survey is an in-depth measurement of car floor and seat cleanliness only.

Rating System PES surveyors evaluate the cleanliness of subway car floors and seats with a four-level rating system in which a score of 1 represents the highest cleanliness rating and a score of 4 represents the lowest. The following definitions are from the New York City Transit’s Passenger Environment Survey, 1st Quarter 1997 and are still in use:

score	PES terminology	PES definition
1	None	Basically dirt free.
2	Light	Occasional "ground-in" spots but generally clean.
3	Moderate	Dingy floor, one or two sticky dry spots.
4	Heavy	Heavy dirt; any opened or spilled food, hazardous (e.g. rolling bottles) or malodorous conditions, sticky wet spots, any seats unusable due to unclean conditions.
The PES notes that "the nature of the dirt (e.g. spilled food, malodorous floor/seat condition, etc.) is also considered in addition to just the volume."*

 
In our study, we used a numbered system identical to the one listed above. However, we modified the terms to describe each rating. We believe our definitions more accurately reflect the meaning behind each score:

score	PES terminology	Straphangers Campaign terminology
1	None	Extraordinarily Clean
2	Light	Clean
3	Moderate	Dirty
4	Heavy	Heavily Dirty

 
Survey

Our project directors trained 28 additional surveyors. During the period October 8, 2002 to January 13, 2003, this team rated exactly 100 subway cars on each of 22 lines*. This is similar to the size of PES samples. We dropped from consideration three lines operated by New York City Transit—the Grand Central-Times Square, Rockaway and Franklin Avenue Shuttles—because they are not rated in the PES survey. Due to similarities in route patterns, observations of the 7- and 7-diamond lines, as well as the Q- and Q-diamond lines, were combined. The measurements were taken both on weekdays and weekends, during rush hours, and during evening and overnight hours.

Analysis of Survey Data

All survey data submitted was visually inspected for error and then coded for entry into a Microsoft Excel 8.0 spreadsheet format. After entering the data, we calculated both by-line and systemwide totals. The percentage of cars rated "dirty" and "heavily dirty" were combined to offer a measurement of the proportion of cars with an unacceptable level of interior dirt. Similarly, "extraordinarily clean" and "clean" rating totals were combined to reflect the proportion of subway cars with an acceptable level of dirt. Systemwide car cleanliness proportion is significant at the .05% confidence level within plus or minus 3%. By-line cleanliness levels are significant at the .05 confidence level within plus or minus 9-11%:

Line	Percent Clean or Extraordinarily Clean	confidence interval plus or minus
1/9	61%	11%
2	55%	11%
3	68%	10%
4	64%	10%
5	68%	10%
6	76%	9%
System Total	59%	3%

Table one notes that in the period between the 2001-2002 and 2002-2003 surveys, car conditions improved significantly systemwide. By-line car conditions improved on nine lines and deteriorated on two lines. Significance of these results was established with a one-tailed Z-test at the .05 level.

Comparison with PES Survey Data

The survey’s finding of improvement mirror the trend in MTA New York City Transit own Passenger Environment Surveys. As of the writing of this report, the transit agency’s ratings for the fourth quarter of 2002 have not been released, so no direct comparison to the Straphangers Campaign survey is possible. In past surveys, New York City Transit has rated subway cars as considerably more clean than has the campaign’s surveys.

Credits

The NYPIRG Straphangers Campaign wishes to thank 28 volunteers who assisted in the survey. They are: Sarah Kogel- Smucker, Theodore Moore, Emily Nella, Meiko Jihan Harris, Martin Markovitz, Jennifer Super, Kellie Roh, Sabrina Matos, Enrique Cepeda, Lynne Asiimwe, Kenneth Sutherland, Tara McManus, Jon Hesser, Aisha Fraites, Asi Klein, Emmanuel Campos, Michael Noda, Delano Aaron, Maria Capitummino, Jack Freedman, Yves Antenor, Monika Kumar, Arelis Hernandez, Michael Hernandez, Andrew Schaaf, Diana Breen, Larissa Madsen, Kenrick Jobe, and Jorge Torres.

_____________________________________________________________________ www.straphangers.org | www.nypirg.org