Second National Communications 5.2 Data collection The following were done in data collection. First we secured monthly/quarterly/annual data on weather parameters from PAG-ASA. For this study, the data for the three pilot sites were available on a daily, quarterly, and annual basis. The definition of a quarter, however, deviates from the standard definition as shown in the table below. PAG-ASA Definition Standard Definition 1st Quarter December, January, February January, February, March 2nd Quarter March, April, May April, May, June 3rd Quarter June, July, August July, August, September 4th Quarter September, October, November October, November, December Second, we secured monthly/quarterly/annual data on diseases from NESSS. For this study, the data provided by NESSS is available by name of hospital, name of patient, date of detection, and patient address. Third, for the correlation analysis between weather parameters and diseases, the data requirements were: number of recorded cases for each disease, observed values of mean temperature (°C), rainfall (mm/day), and relative humidity (%). Fourth, we ensured that the data period for both weather parameters and diseases are the same. Fifth, we generated the number of recorded cases for each disease using the SPSS software. Quarterly data should be consolidated based on the PAG-ASA definition of a quarter to ensure comparability of data. We investigated if there were linear relationships which existed between the variables. For those variables which had a relationship, we would further quantify the relationship to determine how changes in the climate affects disease occurrence. We proceeded to measure the relationship using the Pearson correlation to measure the strength of a linear relationship between two quantitative variables. There is deemed a positive relationship if the generated value is close to +1; a negative relationship if -1 and no relationship if 0. Linear regression quantifies the relationship between variables when the value of one variable is affected by changes in the values of the other variables. It will be good in aiding us assess how well the 31