Food Protection Trends - January/February 2018 - 54

coded with a three-digit identification code unique to each
plant in order to blind responses from participating surveys
to data entry personnel. Participating establishments were
not identified by USDA-FSIS establishment identification
numbers. Survey responses were entered into a database to
facilitate statistical analyses of responses. Data were analyzed
using Microsoft® US Excel® for Mac v.15 (Redmond, CA).
Dillman et al. (13) identified four sources of error in
survey research: coverage error, sampling error, nonresponse
error, and measurement error, each of which should be
addressed and limited to acceptable levels when conducting
survey research. Coverage error occurs when the population
frame fails to include all subjects in the population of interest.
An attempt to reduce coverage error was made by using both
the Texas Association of Meat Processors (TAMP) directory
and the FSIS Meat, Poultry and Egg Product Inspection
Directory. Sampling error occurs when some, but not all,
of the members of the population frame are included in the
study. To reduce sampling error, all processors identified
as Small and Very Small establishments engaging in pork
slaughter and/or processing were invited to participate in the
study. Measurement error is contained within the instrument
and "occurs when respondents are unable and unwilling to
provide accurate answers" (13). Every attempt was made to
incorporate valid, reliable, and unambiguous questions in
the survey instrument by adapting the survey instrument
previously used with other types of NRTE meat products.
Additionally, the instrument was reviewed by a panel of
experts for content and face validity prior to distribution.
Non-response error occurs any time less than 100% response
rates are obtained and can be a threat to external validity (22)

should non-respondents differ from respondents in surveyed
characteristics. A total of 19 usable responses were obtained,
resulting in a 14.5% response rate. Early and late respondents
were compared to address nonresponse error (2, 22, 27).
Late responders were defined as those who responded to the
follow-up questionnaire (22). No significant differences were
found between early and late responders.
RESULTS AND DISCUSSION
Establishments' production and facility characteristics
that potentially impact product microbiological safety
Because of the low response rate, data analysis was limited
to summary statistics. Across all responding facilities, the
mean annual pork production volume was 2.64 million lb.
(range: 20,000 to 10 million lb.). More than two-thirds
(68.4%; 13/19) of establishments reported producing some
type of sausage, including fresh, smoked, skinless, and/or
summer sausage. Over half (57.9%) reported producing
cooked or smoked product(s), while 42.1% (8/19) reported
producing retail pork cuts (Table 1). The most common
products produced by respondents were sausage, cooked
or smoked products, and retail fresh cuts. Food safety
recommendations for Small and Very Small pork harvest and
further processors should thus provide recommendations for
both sets of activities (harvest and fresh products cutting as
well as further processing) for maximal industry utility. The
average facility physical area for responding establishments
(n = 15) was 12,232 + 14,419 ft² (range: 600 to 50,000 ft²),
including non-processing areas (office, access hallways,
product and ingredient/materials storage areas). The average
facility age was 42.4 years (range: 2.0 to 85.0 years; n =

TABLE 1. Fresh and not-ready-to-eat pork products produced by responding small and
very small pork processing establishments (n = 19)a
% Respondents Producing Product Type
Products

Sausage

68.4

Cooked or smoked products

57.9

Retail cuts

42.1

Ham

15.8

Bacon

15.8

Custom pork

10.5

Boudain

5.3

Pork carcasses

5.3

a
Respondents were afforded opportunity to indicate all product classes manufactured; data are presented as percentage of
respondents producing a specific product type.

Food Protection Trends January/February

Table of Contents for the Digital Edition of Food Protection Trends - January/February 2018