Application of Statistical Concepts in the Determination of Weight Variation in Samples

Application of Statistical Concepts in the Determination of Weight Variation in Samples

S.M.D. Claro

National Institute of Geological Sciences, College of Science

University of the Philippines, Diliman, Quezon City, Philippines

Date Submitted: April 23, 2013

Abstract

This experiment aimed to determine the exact weight of a 25-centavo coin through weighing out replicate samples on an analytical balance using the method of weighing by difference. The samples were divided into two data sets and analyzed using statistical concepts for comparison. After applying statistical parameters to the data obtained, outlying values were discarded from the data to ensure the accuracy of the results. By performing the statistical analysis, the experimenter was able to account errors and discrepancies in the experiment. The results showed that data obtained are precise but are not accurate due to some factors. Furthermore, the experiment was done to gain understanding of the statistical concepts applying to analytical chemistry that will be used in future experiments.

RESULTS AND DISCUSSION

After the Q-test was performed, outliers were determined from the two data sets (see Table 1). Q-test was done when dealing with a small number of observations. “In the calculations of mean and standard deviations, the suspect value should be rejected if Qexp is greater than Qtab and should be accepted if Qtab exceeds Qexp.”[1] As for Data Set 1, the highest value was removed from the calculation of mean and standard deviations because Qexp exceeds Qtab. Same goes for Data Set 2. “Rejecting the outliers was only carried out when assuming that the distribution of the population data is normal.” [1]

Table 1. Q-test

Data Set | Suspect Values | Q tab | Q exp | Conclusion |

1 | H: 3.7873 | 0.625 | 0.7876 | rejected |

| L: 3.5371 | 0.625 | 0.1519 | accepted |

2 | H: 3.7873 | 0.466 | 0.7466 | rejected |

| L: 3.5371 | 0.466 | 0.1239 |...