Raw data from the data recording systems (Agilent data acquisition and control unit (DAU)) for beef was loaded into a spreadsheet. The raw temperature data for each sample was extracted from the complete data set adjusted using the pure water calibration point for each individual thermocouple, and placed in a separate spreadsheet headed with sample number and measured pH. The data was then sorted by pH and grouped into low (pHb5.8) and high (pH>6.2) ultimate pH (pHu). The data for each pH range was then averaged and at this point a 43 data point rolling average was applied to all the temperature data to smooth the data. The original data, adjusted for the thermocouple calibrations, was used for the statistical analysis. A Bayesian smoothing technique (Upsdell, 1994) was used to estimate the splines and their least significant intervals (Pb0.05). Ifintervals do not overlap then the splines are significantly different at
the 5% level of significance.