APPLICATION OF STATISTICAL CONCEPTS IN THE DETERMINATION OF WEIGHT VARIATION IN SAMPLES

SALVADOR MARROD M. CRUZ DEPARTMENT OF CHEMICAL ENGINEERING, COLLEGE OF ENGINEERING UNIVERSITY OF THE PHILIPPINES, DILIMAN, QUEZON CITY, PHILIPPINES

ABSTRACT Sampling is the process of obtaining a small portion of a population to represent the population as a whole. Statistics is used to estimate the true value of a certain quantity obtained by using the sample. Statistics is applied in this study in order to obtain the weight of a twenty five centavo coin analytically. In the study, the samples of different sizes were weighed using the analytical balance by the principle of weight difference. The study was performed with 95% confidence level which yielded precise estimates for the true weight of a twenty five centavo coin. The study was able to give a precision of up to 0.03 grams from all of the data sets collected by all groups of researchers.

INTRODUCTION
Analytical chemistry is a branch of chemistry that deals with the identifying and determining the amounts of different components present in a system. Components of a system are analysed by means of conducting experiments. In experiments, a data is being gathered by the researchers in order to achieve a desired quantity or value. In an experiment, errors are present which causes deviations in the data. By means of statistics, the true value of a quantity in a given data can be obtained by using a method called sampling and the errors in the experiment can be estimated. In sampling, researchers will collect different samples of the same analyte (component being analysed) and will collect data in order to obtain an estimate of the true value of a given quantity. The true value can be obtained using the central tendencies namely mean and median. The mean is obtained by getting the average of all the given data (eq.1). The median is the middle value in a data by arranging the data in ascending order and effective in a data with an outlier (a value in a data that is far from other values). Eq.1 while precision describes the closeness of a value with other values present within the data obtained. Accuracy can be obtained by getting the absolute error or the relative error which is just the absolute error divided by the true value (eq.2). Precision can be obtained in different ways it can be obtained by getting the average deviation, standard deviation, variance, relative standard deviation, coefficient of variation or pooled standard deviation if experiment involves more than one data set. But the most commonly used are the standard deviation (eq.3), and pooled standard deviation which both shows the uncertainty of the data. Eq.2
| ( )|

Eq.3. ( ) data sets. (

( )

( )
)

, where ns = number of

In an experiment, the terms accuracy and precision must be taken into consideration. Accuracy describes the closeness of the experimental value obtained to the actual value of the quantity desired

In an experiment the true value of the mean( ) of all values in a given data from the analysis of a sample that may represent whole population lies in the confidence interval (eq.4) which depends on the confidence level of the experiment which increases as the number of data sets increases.

Eq.4 , where the value of t is listed in a table and is dependent on the confidence level of the experiment. For an experiment which contains outliers, the Qtest is done in order to remove the outlier which is commonly found at the extremes when the values of the data are arranged in ascending order and are also called suspect values. The Q-test is done by comparing the Qexp (obtained using eq.5) and Qtab such that if Qtab is greater than Qexp then the suspect value is accepted; otherwise, the suspect value is rejected. Eq.5
| |

under the assumption that determinate errors are absent. The 25 centavo coins were placed in a watch glass for weighing by using a pair of forceps or crucible tongs in order to avoid involving additional components into the coin which may produce a determinate error or an error which can be determined, and an error which can be controlled by proper handling of equipment and careful processing. It is an error which causes a unidirectional effect unto a certain parameter. An example of which is the addition of moisture unto the coin which causes an increase in weight of the coin. Another example is the presence of a foreign contaminant like small objects with a weight able to affect the overall weight of the system. A gross error may also occur or an error which is basically based on serious personal errors, an example of which is the involvement of large amounts of water that could be present in the hand of the researcher upon placing the coin unto the watch glass which sometimes result to the need to repeat the whole experimental process and the use of a new set of samples since the original sample may have deviated significantly from its initial condition. The researchers then weighed the coins in an analytical balance by the principle of weight difference in order to get the weight by removing the coins until only the watch glass is left. The data was then recorded by the researchers. The data recorded by the researchers were subdivided into two sets wherein the first set consists of all the recorded values from the first to the sixth coin and the second set consists of all the recorded values for all of the ten coins weighed. Both sets are listed in table.1 located in the appendix section. Upon obtaining data from the weighing process, the Q test was performed by the researchers by using equation 5 for each of the data sets in order to remove the outliers present in the data (all solutions are located in the appendix section).The outliers are the values which deviate much from the other values in a data due to determinate or gross errors. The Q test was performed also in order to remove values which may be caused by determinate errors which must be absent in order to account for the mean as an estimate of the true mean of the data. And according to both of the data sets recorded by the researchers and of the other groups of researchers, the extreme values were not considered as outliers since they do not deviate much from each other and since all of the values passed all of the individual Q-tests performed for each data set recorded. Since the Q test was not able to remove any of the recorded data, then determinate errors were assumed to be absent when the experiment was performed and statistics can further be used to determine the weight of the coin.

Note: The data set must be arranged in ascending or descending order to obtain the extreme value which is just either the maximum or minimum value in a data set. In the experiment proper, the researchers gathered ten 25 centavo coins that were then weighed using an analytical balance, the purpose of the experiment is to estimate the true weight of the 25 centavo coin by using statistics on two different data sets with different sample size performed at 95% confidence level.

METHODOLOGY
The researchers gathered ten 25 centavo coins of the same year for the experiment proper. The 25 centavo coins were cleansed to remove components that are not part of the coin to avoid additional errors. The 25 centavo coins were then placed in a clean watch glass for weighing. The watch glass along with the coins was placed in an analytical balance and was zeroed by pressing the on tare button. One coin was then removed using a pair of forceps and the weight displaced in the analytical balance was recorded. Then the analytical balance was then zeroed again in order to easily get the weight of the other coins. The process was repeated until no more coins were left in the watch glass for weighing.

RESULTS AND DISCUSSION

In the experiment, the researchers determined the weight of a 25 centavo coin by gathering a sample consisting of ten 25 centavo coins that they analysed statistically using two data sets in order to give an estimate of the true weight of the coin and give an interval where the true mean of the coin is located

Upon performing the Q-test, the researchers were now able to solve for the mean, standard deviation, confidence interval, and pooled standard deviation for the two data sets and are listed in table.2 under the appendix section. The other groups of researchers were also able to do the same and their data are located under the appendix solutions section. The researchers first solved for the mean of the data by using equation 1 which could represent the true mean of a population of data. The mean obtained by the researchers were 3.5884 grams and 3.5973 grams for data sets 1 and 2 respectively, the mean obtained by all of the other researchers are found in the appendix section under the solutions part. The two values were quite close to each other since the coins were produced in the same year and differed only by a small amount from each other as well as the other values in the data sheet attached to this paper because of different sources of errors that may be out of human control which are also called as indeterminate errors or simply random errors, an example of which are the movements of the molecules of molecular differences or something that cannot be noticed easily and something that has a bidirectional effect unto different parameters. Random errors in the experiment performed by the researchers were reduced in the second data set due to the increase in sample size. The Gaussian curve gives the plot of the relative frequency of occurrence of random errors versus the deviation from the mean which is a bell shaped curve on where the maximum height of the curve occurs when errors are most frequent and when it happens, the true mean is obtained at that point. The curve gives a symmetrical view of the distribution of a data. The Gaussian distribution becomes a normal distribution at the time when determinate errors are absent since the normal distribution occurs naturally due to random errors. Due to fluctuations present in the values, the researchers determined the precision of the data by obtaining the standard deviation of the values by using eq.3 in order to determine the uncertainty present within the values due to errors that cannot be determined. Due to the assumption that only random errors were the reason for the fluctuations which is valid since the Q-test was responsible for removing values caused by determinate errors, the standard deviation can be used as an estimate of the population standard deviation of the data which shows the possible range of deviation for a given data. According to the result of the experiment, the researchers obtained a standard deviation of 0.03 grams for both data sets which means that the obtained data is precise up to a sum and difference of 0.03 grams and that the data is uncertain starting from the second decimal place to the right of the decimal point.

From the standard deviation and the mean obtained by the researchers, the confidence interval was then defined in order to give an interval where the true value of the weight of the coin lies. It was computed using equation 5. The confidence interval depends on the confidence level or the percentage of sureness of the researcher that the true mean is in that interval. The confidence interval is also affected by the sample size due to the fact that a larger data set is suffice to show that a closely packed data with a larger number of values ensures the existence of the true value in that interval. For the first data set, the confidence interval obtained was and for the second data set was . The second data set has a lower interval since it involves more values that give a better assurance that the true value of the mean is in a smaller interval where almost all of the other values cluster. Upon obtaining the individual standard deviations of the two data sets, the researchers obtained the standard deviation of the two data sets together by pooling the standard deviations in order to obtain the standard deviation of the two data sets as a whole. The pooled standard deviation can be obtained using eq.3 and by using that equation, the researchers obtained a pooled standard deviation of 0.03 grams which is similar to the individual standard deviations of the two data sets. With the pooled standard deviation, the range of the values for the two data sets can be easily obtained as a whole by using the pooled standard deviation. The result of the data sets 1 and 2 differ from the other data sets collected by other researchers due to different reasons. The most probable reason would be the fact that the coins weighed by different researchers were not made or released the same year as the coins used by the researchers for data sets 1 and 2. Difference in year of production causes deviations since the quantity or amount of material may have been reduced or increased due to different reasons that are beyond the scope of this paper. Other reasons would also be because of the impurities inside the material that cannot be determined by the researchers.

SUMMARY AND CONCLUSIONS

The study was successful because it was able to yield precise results which are primarily caused by random errors that cannot be determined the main probable cause of error. The problem with the study was the differences in the mean value which is due to different factors regarding how the coin was made that made it hard for the researchers to give the interval where the true weight of the coins lie but since the data sets per each group of researchers were obtained from

samples produced in the same year, then the true weight of the coin produced in that year lies within the confidence interval obtained by the researchers. The study could further be improved if the coins by all of the groups of the researchers were produced in the same year and if the researchers will heat the coins in an oven before weighing in order to remove moisture present in the coins to further prevent determinate errors. But overall, the study was successful especially with its precise results.

REFERENCES

[1]Harris, D. C. Quantitative chemical analysis. (8th ed.)(2010). Pg 51-89 [2]Skoog, D. A. Fundamentals of Analytical Chemistry (8th ed.). (2010). Pg 105-169 [3]Walpole, R. E. Probability and Statistics for Engineers and Scientists (9th ed.). (2012). Pg. 1-30

APPENDIX Tables:
Table.1. (Raw data) Data collected by all groups of researchers. The data of group 4 (in orange) are personal data collected by the researchers who created this paper. Legend: Blue = data set 1, light orange = data set 1 of personal data. 1 2 3 4 5 6 7 8 9 10 3.5692 3.5806 3.6069 3.5878 3.6073 3.5803 3.5730 3.5500 3.5873 3.5922 3.6206 3.5407 3.5484 3.6039 3.5860 3.6136 3.5447 3.5755 3.6553 3.6089 3.6033 3.5928 3.5860 3.5722 3.6080 3.5920 3.5590 3.5856 3.6448 3.5960 3.5254 3.5744 3.5993 3.6227 3.6218 3.6464 3.5989 3.5740 3.6255 3.6087 3.5686 3.5720 3.5798 3.5453 3.5900 3.5841 3.5828 3.5901 3.6241 3.6296 3.5966 3.5328 3.5672 3.5985 3.5837 3.5691 3.5463 3.5676 3.6458 3.5893 3.6105 3.5598 3.6093 3.6285 3.5565 3.6254 3.5433 3.5774 3.6740 3.5961 3.5937 3.5178 3.5788 3.5574 3.6118 3.6203 3.5913 3.5608 3.6185 3.6138 3.6461 3.5998 3.6267 3.6241 3.6230 3.5416 3.5226 3.6048 3.6295 3.5645 3.5386 3.5476 3.6042 3.6327 3.5647 3.5887 3.6196 3.5310 3.6577 3.5650 Table.2. Mean, standard deviation, confidence interval, pooled standard deviation obtained from the recorded data by group 4 (personal data). Units of values are in grams. Data Set 1 Data Set 2 Mean 3.5884 3.5973 Standard 0.03 0.03 Deviation Confidence interval Pooled Standard 0.03 Deviation

Solutions: Group 4 (Personal Data): Q-TEST:

Set 1: [3.5453, 3.5722, 3.5878, 3.5985, 3.6039, 3.6227] Set 2: [3.5453, 3.5574, 3.5722, 3.5878, 3.5985, 3.6039, 3.6227, 3.6241, 3.6285, 3.6327] Set 1: Qtab at 95% Confidence Level = 0.625 Extreme values : 3.5453, and 3.6227 Range = 3.6227-3.5453 = 0.0774 Qtest for min value : Since Qexp < Qtab; hence, 3.5453 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6227 is accepted as a part of the data. Set 2: Qtab at 95% Confidence Level = 0.466 Extreme values : 3.5453, and 3.6327
| | | |

Range = 3.6327-3.5453 = 0.0874

|

Qtest for min value : Since Qexp < Qtab; hence, 3.5453 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6327 is accepted as a part of the data.
| |

Mean:
Set 1:

Set 2:

Standard Deviation:
Set 1 ( = 0.0269783617 or 0.03g. Set2: ( ( ) ( ) ( ( ) ( ) ( ) ) ( ) ( ) ( ) ) ) ( ) ( ( ) ) ( ) ( )

= 0.03100406604 or 0.03g.

0.0813971 Relative Standard Deviation

Set 1:

Set 2:

Confidence Limit or interval (95% Confidence):

Set1:
( )( ( )( ) )

Set2:

Pooled Standard Deviation:

( ) ( ) ( ) ( )

Data of other groups (Solutions are the same as those presented for group 4 personal data. All of the answers presented for the other groups were obtained by using a calculator.): Group1:
Set 1: [ 3.5254, 3.5686, 3.5692, 3.5966, 3.6033, 3.6206] Set 2: [ 3.5254, 3.5386, 3.5686, 3.5692, 3.5937, 3.5966, 3.6033, 3.6105, 3.6206, 3.6461]

Q test

Set 1: Qtab at 95% Confidence Level = 0.625 Extreme values : 3.5254, and 3.6206

Range = 3.6206-3.5254=0.0952
|

Qtest for min value : Since Qexp < Qtab; hence, 3.5254 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6206 is accepted as a part of the data. Set 2: Qtab at 95% Confidence Level = 0.466 Extreme values : 3.6461and 3.5254 Range = 3.6461-3.5254= 0.1207 Qtest for min value : Since Qexp < Qtab; hence, 3.5254 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6461 is accepted as a part of the data.
| | | | | |

Mean:

Set 1 : 3.580616667 or 3.5806 grams Set 2: 3.58726 or 3.5873 grams

Standard Deviation

Set 1: 0.03374103832 or 0.03 grams Set 2: 0.03707140857 or 0.04 grams

Relative Standard Deviation:

Set 1: 9.42324 ppt Set 2: 10.334 ppt Set 1: 3.5806 Set 2: 3.5873

Confidence Limit or Interval Pooled Standard Deviation: 0.03591745729 or 0.04 grams Group2: Q test
Set 1: [ 3.5328, 3.5407, 3.5720, 3.5744, 3.5806, 3.5928] Set 2: [3.5178, 3.5328, 3.5407, 3.5476, 3.5598, 3.5720, 3.5744, 3.5806, 3.5928, 3.5998] Set 1: Qtab at 95% Confidence Level = 0.625 Extreme values : 3.5328, and 3.5928 Range = 3.5928-3.5328=0.06 Qtest for min value : Since Qexp < Qtab; hence, 3.5328 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.5928 is accepted as a part of the data. Set 2: Qtab at 95% Confidence Level = 0.466 Extreme values : 3.5998 and 3.5178 Range = 3.5998-3.5178 = 0.082 Qtest for min value : Since Qexp < Qtab; hence, 3.5178 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.5998 is accepted as a part of the data.
| | | | | | | |

0.0354 grams 0.0265 grams

Mean:

Set 1 : 3.56555g. or 3.5656g. Set 2: 3.56183g. or 3.5618g.

Standard Deviation

Set 1: 0.02357572904 or 0.02g. Set 2: 0.02676872346 or 0.03g.

Relative Standard Deviation:

Set 1: 6.6121 ppt. Set 2: 7.5154 ppt.

Confidence Limit or Interval

Set 1: Set 2:

Pooled Standard Deviation: 0.02567399462 g. or 0.03g. Group3: Q test

Set 1: [3.5484, 3.5672, 3.5798, 3.5860, 3.5993, 3.6069] Set 2: [3.5484, 3.5672, 3.5788, 3.5798, 3.5860, 3.5993, 3.6042, 3.6069, 3.6093, 3.6267] Set 1: Qtab at 95% Confidence Level = 0.625 Extreme values : 3.5484, and 3.6069 Range = 3.6069-3.5484=0.0585 Qtest for min value : Since Qexp < Qtab; hence, 3.5484 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6069 is accepted as a part of the data. Set 2: Qtab at 95% Confidence Level = 0.466 Extreme values : 3.5484 and 3.6267 Range = 3.6267-3.5484= 0.0783 Qtest for min value : Since Qexp < Qtab; hence 3.5484 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6267 is accepted as a part of the data.
| | | | | | | |

Mean:

Set 1 : 3.581266667 or 3.5813 grams Set 2: 3.59066 or 3.5907 grams

Standard Deviation

Set 1: 0.02137715291 or 0.02 grams Set 2: 0.02305704038 or 0.02 grams

Relative Standard Deviation:

Set 1: 5.9692ppt Set 2: 6.4214ppt Set 1: 3.5813 Set 2: 3.5907

Confidence Limit or Interval Pooled Standard Deviation: 0.02247150153 or 0.02 grams Group5: Q test
Set 1:[3.5837, 3.5860, 3.5900, 3.6073, 3.6080, 3.6218] Set 2: [ 3.5565, 3.5647, 3.5837, 3.5860, 3.5900, 3.6073, 3.6080, 3.6118, 3.6218, 3.6230] Set 1: Qtab at 95% Confidence Level = 0.625 Extreme values : 3.5837, and 3.6218 Range = 3.6218-3.5837=0.0381 Qtest for min value : Since Qexp < Qtab; hence, 3.5837 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6218 is accepted as a part of the data. Set 2: Qtab at 95% Confidence Level = 0.466
| | | |

0.0224 grams 0.0165 grams

Extreme values : 3.5565 and 3.6230 Range = 3.6230-3.5565= 0.0665 Qtest for min value : Since Qexp < Qtab; hence, 3.5565 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6230 is accepted as a part of the data.
| | | |

Mean:

Set 1 : 3.599466667 or 3.5995 grams Set 2: 3.59528 or 3.5953 grams

Standard Deviation

Set 1: 0.01518231427 or 0.02 grams Set 2: 0.02294441593 or 0.02 grams

Relative Standard Deviation:

Set 1: 4.2179 ppt Set 2: 6.3818 ppt Set 1: 3.5995 Set 2: 3.5953

Confidence Limit or Interval Pooled Standard Deviation: 0.02051224257 or 0.02 grams Group6: Q test
Set 1: [3.5691, 3.5803, 3.5841, 3.5920, 3.6136, 3.6464] Set 2: [ 3.5416, 3.5691, 3.5803, 3.5841, 3.5887, 3.5920, 3.6136, 3.6203, 3.6254, 3.6464] Set 1: Qtab at 95% Confidence Level = 0.625 Extreme values : 3.6464, and 3.5691 Range = 3.6464-3.5691=0.0773 Qtest for min value : Since Qexp < Qtab; hence, 3.5691 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6464 is accepted as a part of the data. Set 2: Qtab at 95% Confidence Level = 0.466 Extreme values : 3.6464and 3.5416 Range = 3.6464-3.5416= 0.1048 Qtest for min value : Since Qexp < Qtab; hence, 3.5416 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6464 is accepted as a part of the data.
| | | | | | | |

0.0159 grams 0.0164 grams

Mean:

Set 1 : 3.597583333 or 3.5976 grams Set 2: 3.59615 or 3.5962 grams

Standard Deviation

Set 1: 0.0281516903 or 0.03 grams Set 2: 0.03063205474 or 0.03 grams

Relative Standard Deviation:

Set 1: 7.8252 ppt Set 2: 8.5180 ppt Set 1: 3.5976 Set 2: 3.5962

Confidence Limit or Interval Pooled Standard Deviation: 0.02976994329 or 0.03 grams

0.0295 grams 0.0219 grams

Group7: Q test

Set 1: [3.5447, 3.5463, 3.5590, 3.5730, 3.5828, 3.5989] Set 2: [3.5226, 3.5433, 3.5447, 3.5463, 3.5590, 3.5730, 3.5828, 3.5913, 3.5989, 3.6196] Set 1: Qtab at 95% Confidence Level = 0.625 Extreme values : 3.5989, and 3.5447 Range = 3.5989-3.5447=0.0542 Qtest for min value : Since Qexp < Qtab; hence, 3.5447 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.5989 is accepted as a part of the data. Set 2: Qtab at 95% Confidence Level = 0.466 Extreme values : 3.6196and 3.5226 Range = 3.6196-3.5226 = 0.097 Qtest for min value : Since Qexp < Qtab; hence, 3.5226 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6196 is accepted as a part of the data.
| | | | | | | |

Mean:

Set 1 : 3.56745 or 3.5675 grams Set 2: 3.56815 or 3.5682 grams

Standard Deviation

Set 1: 0.02140941382 or 0.02 grams Set 2: 0.03012954438 or 0.03 grams

Relative Standard Deviation:

Set 1: 6.0013 ppt Set 2: 8.4440 ppt Set 1: 3.5675 Set 2: 3.5682

Confidence Limit or Interval Pooled Standard Deviation: 0.02733642259 or 0.03 grams Group8:
Set 1: [ 3.5500, 3.5676, 3.5740, 3.5755, 3.5856, 3.5901] Set 2: [ 3.5310, 3.5500, 3.5608, 3.5676, 3.5740, 3.5755, 3.5774, 3.5856, 3.5901, 3.6048] 0.0225 grams 0.0215 grams

Q test

Set 1: Qtab at 95% Confidence Level = 0.625 Extreme values : 3.5901, and 3.5500 Range = 3.5901-3.5500=0.0401 Qtest for min value : Since Qexp < Qtab; hence, 3.5500 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.5901 is accepted as a part of the data. Set 2: Qtab at 95% Confidence Level = 0.466 Extreme values : 3.6048 and 3.5178 Range = 3.6048-3.5310 = 0.0738 Qtest for min value : Since Qexp < Qtab; hence, 3.5310 is accepted as a part of the data.
| | | | | |

Qtest for max value = Since Qexp < Qtab; hence, 3.6048 is accepted as a part of the data.

Mean:

Set 1 : 3.5738 grams Set 2: 3.57168 or 3.5717 grams

Standard Deviation

Set 1: 0.01423193592 or 0.01 grams Set 2: 0.02093003159 or 0.02 grams

Relative Standard Deviation:

Set 1: 3.9823 ppt Set 2: 5.8600 ppt Set 1: 3.5738 Set 2: 3.5717

Confidence Limit or Interval Pooled Standard Deviation: 0.01881362728 or 0.02 grams Group9: Q test
Set 1: [3.5873, 3.6241, 3.6255, 3.6448, 3.6458, 3.6553] Set 2: [ 3.5873, 3.6185, 3.6241, 3.6255, 3.6295, 3.6448, 3.6458, 3.6553, 3.6577, 3.6740] Set 1: Qtab at 95% Confidence Level = 0.625 Extreme values : 3.6553, and 3.5873 Range = 3.6553-3.5873=0.068 Qtest for min value : Since Qexp < Qtab; hence, 3.5873 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6553 is accepted as a part of the data. Set 2: Qtab at 95% Confidence Level = 0.466 Extreme values : 3.6740 and 3.5873 Range = 3.6740-3.5873= 0.0867 Qtest for min value : Since Qexp < Qtab; hence, 3.5873 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6740 is accepted as a part of the data.
| | | | | | | |

0.0149 grams 0.0150 grams

Mean:

Set 1 : 3.630466667 or 3.6305 grams Set 2: 3.63625 or 3.6362 grams

Standard Deviation

Set 1: 0.02443691197 or 0.02 grams Set 2: 0.02456801217 or 0.02 grams

Relative Standard Deviation:

Set 1: 6.7311 ppt Set 2: 6.7564 ppt Set 1: 3.6305 Set 2: 3.6362

Confidence Limit or Interval Pooled Standard Deviation: 0.02452127113 or 0.02 grams Group10:
Set 1: [ 3.5893, 3.5922, 3.5960, 3.6087, 3.6089, 3.6296] Set 2: [ 3.5645, 3.5650, 3.5893, 3.5922, 3.5960, 3.5961, 3.6087, 3.6089, 3.6138, 3.6296] 0.0256 grams 0.0176 grams

Q test

Set 1: Qtab at 95% Confidence Level = 0.625 Extreme values : 3.6296, and 3.5893 Range = 3.6296-3.5893=0.0403 Qtest for min value : Since Qexp < Qtab; hence, 3.5893 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6296 is accepted as a part of the data. Set 2: Qtab at 95% Confidence Level = 0.466 Extreme values : 3.6296 and 3.5645 Range = 3.6296-3.5645= 0.0651 Qtest for min value : Since Qexp < Qtab; hence, 3.5645 is accepted as a part of the data. Qtest for max value = Since Qexp < Qtab; hence, 3.6296 is accepted as a part of the data.
| | | | | | | |

Mean:

Set 1 : 3.604116667 or 3.6041 grams Set 2: 3.59641 or 3.5964 grams

Standard Deviation

Set 1: 0.01497136155 or 0.01 grams Set 2: 0.02050585011 or 0.02 grams

Relative Standard Deviation:

Set 1: 4.1540 ppt Set 2: 5.7018 ppt

Confidence Limit or Interval

Set 1: 3.6041 Set 2: 3.5964 0.0157 grams 0.0147 grams

Pooled Standard Deviation: 0.01871805342 or 0.02 grams Pooled Standard Deviation for all groups: (
( ) )

= 0.03 grams