Statistical Parameters of the Company ABC - Essay Example

 Scope of Work The company ABC has provided quarterly sales data for years from the year 2000 through 2004. The data set consists of 20 values. This assignment uses the values of the data set to study statistics of the past performance of the company and predict the future quarterly sales for years 2005 and 2006. In studying the past performance, the assignment compares the past quarterly sales through statistical parameters such as mean, median, quartiles, and bar chart. The future quarterly sales are predicted using regression analysis and moving average method. Quarterly Sales Comparison. Table 1 displays the past quarterly sales values for years from 2000 to 2004. The analysis focuses on sales in each quarter of years from 2000 to 2004. The data structure allows viewing dynamic of sales in a quarter through all four years as well as trend through four quarters of each year. Figure 2 shows that sales in q1, q2, q3, and q4 grew in every consecutive year with respect to the previous year. On the other hand, Figure 1 displays that sales in the second quarter in every year decreased with respect to the first quarter; however, it kept growing in q3, and q4. This implies the seasonal influence on sales of this product. Nevertheless, the general trend of sales per year demonstrates growth, which is also confirmed by the mean and median values displayed in Table 1 and Figure 3. Figure 3. Mean and Median growth per year due to sales growth Percentile and Quartile Comparison of Sales. The data set consists of 20 values. The company achieved 25 percentile in the fourth quarter of 2000; 50 percentile in the second quarter of year 2003, and 75 percentile in the third quarter of 2003 (see Table 2). The study reveals that it took 14 quarters to achieve first 50 percentile while the company achieved the second 50 percentile in 6 quarters. The growth should be considered as fast starting from the year 2003. In this year, the company achieved 17% growth in the third quarter with respect to second quarter. Summary of Data Set. Shown below histogram of the values of the data set along with the position of the mean, first quartile, second quartile, and third quartile. Figure 4. Data set summary Percentile 10 25 50 75 95 Quartile Q1 Q2 Q3   Values 683.48 890.58 1211.9 1671.53 2127.46 Date q1 / 2000 q4 / 2000 q2 / 2003 q1 / 2004 q3 / 2004 Table 2. Percentile and Quartile values Sales Prediction The sample data set consists of 30 values and it shows a trend of sales through five years. The aforementioned section provided an analysis of past trend, which are displayed in tables and graphs. This section focuses on the evaluation of quarterly sales for the next two years based on the history of previous sales. The future sales prediction is performed using regression analysis and time series technique. Regression Analysis. Regression analysis prediction is based on the interdependency of two variables; one is independent and the other is dependent. In this study, sales are the dependent variables that follow a seasonal trend, which is displayed in Figure 5. The seasonal trend in this analysis is considered as independent variable. The relation is presented through the straight-line equation Y = a + bX, where X represents the season and Y represent the sales. The dependent variables receive values from 1 through 28, where the values, from 1 through 20, display past sales, and from 21 through 28 displays the future values. The intercept “a” and the slope “b” are calculated using the past data; these values are used to predict the future sales. The results are shown in Table 3 and Figure 6. Figure 6. Sales prediction using Regress analysis Time Series Technique. Time series technique studies a history of sales through time; once the pattern is identified the future is projected according to that pattern (“Forecasting Theory”). This study uses moving average method to forecast 2005 and 2006 quarterly sales using time series technique. The method uses the following formula F t + 1 = (St + St-1 + St-2+ …….+ S t-N-1) / N; where: F t + 1 = Forecast for the period t+1; St-1 = Sales for the period t – 1, and N = Number of period in moving average Thus; A three-period moving will be F t+1 = (S t + S t-1+ S t-2) / 3 A four-period moving will be F t+1 = (S t + S t-1+ S t-2 + S t-3) / 4. This study conducted an analysis using 3-day, 4-day, and 5-day moving averages. The purpose is to locate the best suitable curve for forecasting sales. Table 4 and Figure 7 display the results of the calculations. The three curves were compared visually against the curve that represents the original data set. Though all three curves display a trend, however; the graphical observation demonstrates that the 4-day average curve is the best fit to the original data. The forecast sales for years 2005 and 2006 are accepted as follows: Moving Average Method q1 q2 q3 q4 2005 2175 2192 2328 2334 2006 2257 2278 2299 2292 Regression Analysis q1 q2 q3 q4 2005 2320 2413 2506 2598 2006 2691 2783 2876 2969 Work Cited “Forecasting Theory.” utexas.edu. The University of Texas at Austin, n.d. Web. 02 July 2013. http://www.me.utexas.edu/~jensen/ORMM/omie/operation/unit/forecast/moving_avg.html Read More