ABSTRACT

Contaminated and consumed water can be the main cause of foodborne diseases. Therefore,
water needs to be examined bacteriologically as a water quality test. Each examination
method needs to be tested for the validity of the method, such as testing sensitivity and
specificity. This research aims to determine the sensitivity and specificity values of the
miniaturized Most Probable Number method against the Total Plate Number method. This
type of research is quantitative with a pre-experimental research design. The samples used
were bottled drinking water, refillable drinking water, and iced drinks sold around the
Pengasinan area. The samples used in this research were 28 samples, namely 13 positive
samples and 15 negative samples. The data analysis used was the Cohen's Kappa test. The
results were obtained based on the Cohen's Kappa statistical test with a sample of 28. The
mMPN kit was found to have a sensitivity value of 84.6% and a specificity value of 100%.
Bacteriological test results showed that 10 out of 12 samples (83.33%) of iced tea samples
were contaminated with bacterial counts >100 CFU/ml, 3 out of 7 samples (42.85%) of refill
drinking water samples were contaminated with bacterial counts >100 CFU/ml. ml, and 6
samples (100%) of bottled drinking water were not contaminated with bacterial counts