Statistical methods for business decision-making and data analysis

Statistics can be described as the study of the collection, interpretation, presentation, description and analysis of data. In particular, statistics in business taught us to apply a variety of statistical methods in a business context to facilitate evidence-based decision making and to provide answers to business related questions. The overarching objective of the course was to analyze each of the course elements in detail and at the same time learn how to use this knowledge to describe data and make informed decisions based on well reasoned statistical arguments. The course elements learnt include: descriptive statistics, inferential statistics, hypothesis development and testing, selections of appropriate statistical tests, and evaluating statistical results. All the elements enabled us to make inferences about a given population from sample data, to perform statistical analyses and to interpret different results. Thus, all the knowledge gained from the statistics course is applicable in real life situations and will also be useful in solving a variety of analytical problems in our future careers. This text takes a look at the five elements of the business statistical course in detail and evaluates their applicability in day-to-day operations and different careers.

Descriptive statistics

Descriptive statistics are used to describe or summarize data in a meaningful way so that patterns can emerge from the data. Numeric values such as range, standard deviation, mean, mode and median are used to make descriptions about the main characteristics of data in a study. They, however, fail to make conclusions beyond analyzed data or conclude on any hypothesis that had been made about the data (Wasserman, 2004). Descriptive statistics can be applied in a variety of situations. For instance, they can be used to describe findings in business related research, to provide insight on business trends, and to explain deviations from expected levels of performance. Descriptive analysis methods can also be used to explain trends followed by stocks that are traded on a financial market and to explain fluctuations in currency in relation to international trade. The knowledge gained can support a career in the U.S. Census Bureau where descriptive characteristics are used to indicate average household sizes, employment rates, pa-capita income and gender and ethic breakdowns. The methods are also useful for people who conduct market research and offer financial services.

Inferential statistics

Inferential statistics makes use of probabilistic techniques to analyze sample information from a known part of a population to improve knowledge about the unknown whole. More specifically, techniques learnt in inferential statistics allowed us to use samples to make informed generalizations about the populations from which the samples were drawn. Downing and Clark (2010) state that the two methods applied in inferential statistics are the testing of statistical hypotheses and the estimation of parameters. The methods incorporate sampling errors that arise when a sample fails to represent the population perfectly. Inferential statistics are applied in daily managerial decision making, product promotion surveys, marketing and research and competitor analysis. It can also be used in the calculation of the consumer price index (CPI) and the gross domestic product (GDP) for government use. Careers such as professional statisticians and economists, as well as in credit rating agencies and bureaus of statistics, where extensive experiments are required make use of this discipline when in need of information to reach conclusions that involve large populations (Downing and Clark, 2010).

Hypothesis development and testing

Hypothesis testing, also referred to as significance testing, is used when attempting to answer hypotheses or research questions that are set. It provides a basis for evaluating ideas that are developed and testing them to find out whether they are true or false. It also makes use of the sampling technique, where a researcher or statistician may seek to establish whether differences identified in a sample exist in the entire population. Wasserman (2004) explains that during data analysis, hypothesis test are stated in terms of one condition that the statistician or researcher doubts, and one that they believe. The null hypothesis refers to the statement that will be tested often denoted by 'H0', while the alternative hypothesis 'H1' refers to the one that will be accepted once the former is rejected. Hypothesis testing, therefore, starts with making a statement about the value of a parameter, then goes on to establish whether the statement is true or false. Hypothesis testing is useful in real life situations where decision making involves testing various ideas. These include introducing a new product into the market, investigating the effects of a new drug, or even determining the best teaching methods to be applied in an institution of higher learning. Furthermore, careers in different fields of investments, statistical research, and bureaus of statistics, among many others, make use of hypothesis testing.

Selection of appropriate statistical tests

Statistical tests provide an individual with mechanisms for making quantitative decisions about a particular process. The tests determine whether there is sufficient evidence to reject a null hypothesis. Test statistics are the quantities taken from samples and used as the basis of testing a null hypothesis. Significance levels are also taken into consideration because they incorporate the probability of certain type of errors. Statistical tests include: the F-test, Chi-square test, and Levene test for measuring variability; runs tests and autocorrelation tests for measuring randomness; and measures for skewness and Kurtosis (Anderson, Sweeney and Williams, 2013). All statistical tests can be applied in the day-to-day operations of a variety of careers to address uncertainties of sample estimates.

Evaluating statistical results