Data Collection and Analysis Paper - Statistics Project Example

Statistics Project, Statistics Data Collection and Analysis Paper Assignment METHOD OF DATA COLLECTION This project involves the collection and analysis of data on the soldiers in the SRP (Soldiers Readiness Process) in the US Army. The SRP site is designed for the demobilization of the soldiers for the purpose of expansion of the military power. The proposed test to reduce the turnaround time for each soldier is done by sampling different number of soldiers between in multiples of 5 and measuring the time taken to deliver the results. The maximum number of soldiers is 300. The method used in collecting the data is by direct observation, rather than questionnaire, secondary data and interview methods. Data Measurement Plan The main data measured in this exercise is the time factor against the number of soldiers in each test. The plan for measurement is the preparation of measuring equipment such as an alarm clock and a stop watch. It is hypothesized that the turnaround time for each group reduces with the increase in the number of soldiers in the group. It implies that many soldiers take shorter time than a few soldiers do. This adds value to the demobilization through the power of leverage. Consequently, the data stratification tree is demonstrated in figure 1 below. Figure 1: Data Stratification Tree The tree is wide at the bottom but it diminishes upwards as the number of soldiers reduces. It implies that more time is required with a few soldiers, but as the number grows, less turnaround time is observed. The measurement of time gives the data as shown in table 1 below. Table 1: Time Data Measurement Number of Soldiers In a Group Turn Around Time (Minutes) 10 75.00 20 37.50 30 25.00 40 18.75 50 15.00 60 12.50 70 10.71 80 9.38 90 8.33 100 7.50 110 6.82 120 6.25 130 5.77 140 5.36 150 5.00 160 4.69 170 4.41 180 4.17 190 3.95 200 3.75 210 3.57 220 3.41 230 3.26 240 3.13 250 3.00 260 2.88 270 2.78 280 2.68 290 2.59 300 2.50 Figure 2: Time Measured Against the Number of Soldiers Actual Measurement of Data The types of information measure is the time taken by the number of soldiers in the groupings. It is intended to measure the optimal processing time in the demobilization and to reduce the congestion and confusion among the soldiers. The processing is observed to reach a constant limit at the level of 180 soldiers. It means that the best result of turnaround time was obtained by splitting the soldiers into two groups, one group having 180 soldiers and the second group, 120 soldiers. There is a clear definition for the input and output in this exercise, whereby, the number of soldiers in the groups is the input, while the time measure taken in carrying out the given SRP exercise is the output. The data collected was continuous (with fractions). However, the input data was discrete. The baseline measurement of the input data was 90 minutes, which an individual soldier takes to complete the SRP process. The major objective is to reduce the time by between 33.0% to 60.00 minutes. There is need for development of standard operational procedures for the home units as well as Fort Sill to uphold the efficiency of the SRP processes. Originality of the Data The data collected was original primary data collected from observation. It was not a secondary data because the SPR is a practical exercise, not a historical account of the past events. The data collected were 30 records of data because each group had 10 respondents. This was to reduce the number of tests and improve on the processing time on the data collected. Any number less than 5 would require more than 30 tests and measurements, hence, generating more than enough data. Secondly, any number more than 10 would generate very few records and opportunities for testing. This would lead to insufficient data that eventually leads to high error rate and unreliable output. In this regard, the ideal sample size is using the sample-size equation is calculated as shown below: S = (T) / d Where T is the maximum number of soldiers (T = 300) and d is the gap between each sample group, which in this case is 10. The sample size is obtained as: S = 300 / 10 S = 30 The risk given in the chosen sample is that the selection of soldiers for the groups may not be random; hence there may be two remarkably different and unrealistic outcomes. For a realistic result, there is need to use a random selection of the soldiers for each of the groups. Method of Data Collection The data was collected through direct observation of the practical SPR activities. The soldiers were engaged in the SRC processes while the stop watch was running. Immediately the exercise is declared complete, the stop watch is stopped and the time is recorded, before taking a subsequent group. The measurement error could arise from the rounding off of seconds into the nearest minutes. For example, all the data collected measured the time in two decimal places. This easily hides the round off errors. In order to measure the amount of error hidden in the calculation, there is need to get the actual measurement then subtract the exact measurement taken to get the error. The data is thus presented as seen in table 2 below: Table 2: Error Calculation Soldiers Actual Turn Around Time (Minutes) Round Off Turn Around time Error 10 75 75 0 20 37.5 37.5 0 30 25 25 0 40 18.75 18.75 0 50 15 15 0 60 12.5 12.5 0 70 10.7143 10.71 -0.0043 80 9.375 9.38 0.005 90 8.3333 8.33 -0.0033 100 7.5 7.5 0 110 6.8182 6.82 0.0018 120 6.25 6.25 0 130 5.7692 5.77 0.0008 140 5.3571 5.36 0.0029 150 5 5 0 160 4.6875 4.69 0.0025 170 4.4118 4.41 -0.0018 180 4.1667 4.17 0.0033 190 3.9474 3.95 0.0026 200 3.75 3.75 0 210 3.5714 3.57 -0.0014 220 3.4091 3.41 0.0009 230 3.2609 3.26 -0.0009 240 3.125 3.13 0.005 250 3 3 0 260 2.8846 2.88 -0.0046 270 2.7778 2.78 0.0022 280 2.6786 2.68 0.0014 290 2.5862 2.59 0.0038 300 2.5 2.5 0 Total Error 0.0159 In order to minimize the Error, the measurements can be taken in higher precision values. For example, instead of rounding off to the nearest tenths, we can round off to the nearest the thousandths. This means we will have 4 decimal places instead of two, giving more accurate measurements. DATA ANALYSIS TOOLS The SQL for the old process is the update information that each soldier spends 90 minutes in the SRC process. In the new process however, there is leverage and the soldiers spend an average of 9.98 minutes. The tools for measuring data are classified into two categories, software and hardware. As previously mentioned, in the hardware category, this project required a stop watch and an alarm clock for timing the SRC exercise among the soldiers. In the software category, this project required an excel spreadsheet to apply the computation and analysis of the data. There are three techniques used in this analysis including the measures of central tendency, measures of dispersal and error analysis. This is to verify the consistency of the behavior of the random distribution. The measures of central tendency are presented in the table 3 below: Table 2: Measures of Central Tendency Mean 9.98747 median 4.84375 Mode #N/A Additionally, the measures of dispersal are presented in table 4 below: Table 2: Measures of Central Tendency Standard Deviation 14.47454118 Variance 209.5123425 From the data, the time spent by the soldiers reduced significantly from 90 minutes to less than 10. However, the computed mean of the time spent deviates from the actual mean to the scale of 14.47. CONCLUSION The error analysis exposes the effect of measurement and the treatment of the measurement from which this project proposes an error reduction, by increasing the number of decimal places used in the measurement. The proposal was not implemented, but it was tested and confirmed to be reducing the errors from 1.59% to 0.68%. The lesson learnt in the process is that leverage increases efficiency and accuracy of performance. The old process is not in control of the SRC results. However, the new process is fully in control and is influential on the mobilization and demobilization of soldiers.. Read More