The Application of Business Data Analysis to make Decisions
) Calculate at least 2 measures of central tendency and at least 2 measures of variation (from 2 different countries) for
your Continuous Variable sample data. Explain the results and meaning of these measures in your report. Include a summary of the data in your report using a graph or table. Use Excel for this task.
2) Explain what your Discrete Random Variable is and what probability distribution can be used to describe it. Give some examples and interpretation of the usage of the probability distribution function and outline its business relevance.
3) Either do 3.a or 3.b
a. Calculate and interpret a confidence interval estimate of the Continuous Random
Variable. Choose an appropriate confidence level for your interval.
b. Set up and conduct a hypothesis test relevant to the continuous random variable.
Choose an appropriate level of significance for your test.
4) Show how Regression techniques can be used to extrapolate or interpolate values for your
dependant variable.
Introduction
Nowadays as business environments continue to become more dynamic and competitive locally, regionally and internationally, the application of data analysis and statistics in business as an aid to decision making has become inevitable (Davenport & Harris, 2007). This is attributable to the fact that, decisions in contemporary businesses play an essential role in making sure that enterprises are strategically positioned in the market through management and marketing decisions that are informed by business intelligence analytics (Davenport et al., 2009). Bartlett (2013) notes that business data analysis applications using tools and techniques that are technologically advanced such as business data analysis software have become an integral part of decision making in modern day businesses. Furthermore, business data analysis involves an iterative and methodical exploration of data from an organization with emphasis on statistical analysis in order to make sure that their decision making process is data-driven (Davenport & Harris, 2005). This management report aims to provide a meaningful interpretation of business data analysis (about ObviEnce, LLC sales and marketing data) to the business manager in order to demonstrate its use in decision making. Two variables are used i.e. one discrete random variable (number of new orders) and one continuous random variable (marketing time in months).
Summary and explanation of descriptive statistics
The business data considered is summarized using descriptive statistics as well as a presentation technique. For instance, the descriptive statistics used were measures of central tendency as well as variation whereby the former measures are mean and median while latter measures are standard deviation and variance. These descriptive statistics were calculated for both variables i.e. discrete random variable and continuous random variable. As a result, Table 1 below shows, the business data which was considered in the data analysis using the Ms Excel program together with the subsequent descriptive statistics, both in terms of measure of central tendency as well as measure of variation. Descriptive statistics play an essential role in making sure that a role data is analyzed and presented in a manner that is easily understandable while at the same time making sure that the meaning of the data can also be easily interpreted.
The mean for the data presented in this report is 1760 and 13.3 for the number of new orders (discrete random variable) and marketing time in months (continuous random variable) respectively. In addition, the median for the data presented in this report is 1800 and 14 for the number of new orders (discrete random variable) and marketing time in months (continuous random variable) respectively. The mean shows the average of a data set while median is the middle data point in an ordered data set (Davenport, 2006). From, the descriptive statistics shown in Table 1 below for both variables indicate that the mean and median are very close meaning that the data concerning the impact of marketing on sales is normally distributed, which implies that more data is located in the middle.
On the other hand, the measures of variation were also determined using Ms Excel and the standard deviation of the data presented in this report is 874.2752 and 7.220507 for the number of new orders (discrete random variable) and marketing time in months (continuous random variable) respectively. Furthermore, the median for the data presented in this report is 764357.1 and 52.13571 for the number of new orders (discrete random variable) and marketing time in months (continuous random variable) respectively. The measures of variation show how individual points in a data set are sparsely distributed, and also the extent to which they have deviated from the mean. Higher standard deviation and variance indicate significant deviation of individual data points in a data set from its mean (Stubbs, 2011). For instance, the standard deviation and variance for both variables are significantly varied from the mean. This means that the marketing of the company had impact on sales of the company.
Table 1: Descriptive Statistics
The data can also be presented using a line graph created using Ms Excel as illustrated in the Figure 1 shown below.
Figure 1: A line graph of the data
The discrete random variable used in this data is the number of new orders which are determined by the length of marketing in the number of months. This is mainly because the longer the marketing time, the higher the number of new orders. Discrete probability distribution can be used to describe this variable since the number of orders must be a definite and not partial meaning it, must be represented by a specific number. An example is the tossing of coin whereby you must either get a head or tail meaning after a particular number of tosses there must be a discrete number of heads and tails.
Inferential Statistics
Normal distribution is a symmetric curve consisting of one central peak which is equivalent to mean of the data. Bell-shaped is the word used to describe the curve whereby the graph falls off in an even manner on both sides of the mean and 50% of the distribution is located on either side of the mean (Stubbs, 2011). The use of Normal Distribution in sampling distributions is done because they usually show naturally occurring phenomena which are determined through sampling, in order to get unbiased statistical tests (Davenport & Harris, 2007). Inferential statistics are based on inferring from the sample data in order to make generalizations about the populations which enable conclusions to be made (Davenport et al., 2009).
The continuous random variable used in the data is the length of time (in months) taken by the company in marketing. This is attributed to the fact that time can be presented continuously meaning that a range of time can be used to determine the discrete random variable (Davenport & Harris, 2007). Sampling was done by counting the number of new orders that were placed within a specific number of months in marketing. A confidence interval is the range in which a certain trial or measurement falls in within a specified probability (Stubbs, 2011).
Construction of a confidence interval is first done by identifying the test statistic (e.g., sample proportion and sample mean) subsequent to selection of the confidence level which may be 90%, 95% or 99%. The margin error is then calculated, followed by specifying of the confidence interval. With a sample size of 15 (df=14), mean of 13.3, and standard deviation of 7.220507 confidence interval of the continuous random variable at the 99% confidence level is:
The above confidence interval results show that 99% of the intervals will include the test statistic (mean).
Regression techniques can be used to extrapolate values for the dependent variable by developing with a constant meaning when the independent variable is known then the dependent variable can be determined.
Summary and Conclusion
In summary, the descriptive statistics show that the data values are widely located around the mean and there was a positive correlation between the time take in marketing and the volume of sales in terms of new orders. Inferential statistics also show a considerably wide confidence interval. In conclusion, it has been seen that it is possible use data analysis to make important business decisions.
References
Bartlett, R. (2013). A Practitioner’s Guide to Business Analytics: Using Data Analysis Tools to Improve Your Organization’s Decision Making and Strategy. New York, NY: McGraw-Hill.
Davenport, T. H. (2006). Competing on Analytics. Boston, MA: Harvard Business School Press.
Davenport, T. H., & Harris, J. G. (2005). Automated Decision Making Comes of Age. Boston, MA: MIT Sloan Management Review.
Davenport, T. H., & Harris, J. G. (2007). Competing on analytics: The new science of winning. Boston, MA: Harvard Business School Press.
Davenport, T. H., Harris, J. G., De Long, D. W., & Jacobson, A. L. (2009). Data to Knowledge to Results: Building an Analytic Capability. California Management Review, 43 (2), 117–138.
Stubbs, E. (2011). The Value of Business Analytics. Hoboken, NJ: John Wiley & Sons.
