# Statistics refers to a study of numerical figures. The classification of statistics is wide depending on the discipline of study. People rarely attach statistics beyond just figures.

Clinical Case Assessment

Author’s Name

Affiliated Institution

Statistics and Work

Statistics refers to a study of numerical figures. The classification of statistics is wide depending on the discipline of study. People rarely attach statistics beyond just figures. Interpretation of those figures requires statistical skills which are owned by a few who have majored in statistics. There are various ways of classification of statistics as discussed below

Statistics and Marketing

In marketing, every organization has a role in maintaining relevance to the marketing dynamics by obtaining raw data from the field and then analyzing this data into useful forms that can be applied to outwit a competitor. Market segmentation is classified into the following categories that are used frequently at Songhyun Academy where I work (Bhasin, 2016)

a) Demographic Segmentation

In this case, data is classified into variables consisting of biodata of the target group. The variables include age, sex, income, occupation, religion, socioeconomic status among others. Various products from companies target different customers. For instance, socioeconomic status of customers is important when designing a product. There are those products that target the poor while others target the rich. Carrying out research to analyze such a data is a key to maintaining relevance. For example if a new company has a new book with heading “My life in the 1980s” the target population in this case are those born in the 1980s hence the company will be required to carry out a quantitative research to identify where most of the population with that age group is located to determine the market demand.

b) Behavioral Segmentation

Behavioral segmentation involves analysis of data by using attributes of customers. Data is collected from the target about their attributes to a product and then the attributes are incorporated into the product design. For instance, the introduction of basic computing skills at Songhyun Academy was done to attract more students into the school as more parents attributed the quality of education to exposure of children to basic computing skills. In behavioral segmentation for example, in the food industry, attributes of customers preference are fed into a computer which makes regressions on the attributes of food that make customers grave for more such as addition of more sugar. Quantitative statistics is required to determine which type of food additive and quantity increase graving.

c) Psychoanalytical Segmentation

This is similar to behavior segmentation only that it includes the psychological aspect of the target customers.

d) Geographical Segmentation

Different geographical locations have different market needs. For example in terms of clothing, the types of clothes worn in Polar Regions are different than the one worn in deserts due to variations in climatic conditions.

Pattern Classification

Pattern classification utilizes various techniques to classify data using the aid of computer installed programs. Some of the techniques of computer-aided classifications are as follows

a) Linear Discriminant analysis

Data is first classified using a default linear discriminant analysis and the results are plotted on a graph in which the X-axis contains an independent variable while the Y-axis contains dependent variable (Bamparopoulos, 2012). One product may contain different features from which customers do choose. For instance, soap has features such as weight, size, color, and smell from which people choose. Linear discriminant analyzes these features to get the best combination of qualities for soap for most customers.

Student’s performance at end of the education system is determined by many other variables that can be used to predict student performance using a Quadratic discriminant analysis with SPSS software. The polynomial form ax2 + bx + c=0 can be turned into a quadratic discriminant formula which can be used to predict the performance of students at end of program as follows: x=(-b±√(b^2-4ac))/2a whereby x is the predictable grade.

For example in Songhyun Academy where I work mathematics has a higher failure rate compared to other subjects. Predicting those students who are at risk of failing is determined by using variables that have been confirmed to have an influence on performance in mathematics such as gender, race, and class size. We group this data into the three categories and then substitute the information with the quadratic discriminant formula to identify those students that fail and offer them extra assistance to predict the failure rates. The constant in our test is the minimum score requirement by the national examination council.

b) Bayesian Classifier

The Bayesian classifier is a technique for finding the probability of an event occurring. It is a subjective measurement that tries to measure the likelihood an event occurring. Unlike frequency which measures the number of times an event occurs, it measures the likelihood of that event occurring. It is calculated using Bayesian inference.

Clustering of Statistics

Clustering of statistics is an unsupervised grouping of data. The classification can be hierarchical algorithms, density based, grid-based, model-based clustering or partitioning algorithms (Biehl, 2009).

Finance statistics

Statistical finance arose from a branch of physics and it involves the use of physics to analyze the statistical data in finance. The methods of statistical analysis are as follows

Mean

This is the sum of all sets of data divided by the number of items that are on that list.

Standard deviation

This refers to the spread of data away from the mean

Regression

This refers to then the relationship between variables

Finance can also be categorized into the following subcategories into two major categories namely qualitative and quantitative finance

Covariance matrices

Variance and covariance are closely related to each other in that they both measure the spread of dimensions of a set of data. Dimensions refer to variables for example time, age, height, and others. Variance measure the spread of data from the central mean. For example, if the mean for exam of 10 students in a mathematical test is 53%, variance measures how many scored above 53% and below 53%. Covariance is a measure of two dimensions to find out if there is any difference between each other. In the variance we used one dimension for example number of students s as our reference while in the covalence, for example, a comparison can be made between number of hours a student read and the scores away from the mean.

If 3 variables are used, for example x, y and z, they can be presented a matrix as shown below ■( cov(x,x)&cov(x,y)&cov(x,z)@C=cov(y,x)&cov(y,y)&cov(y,z)@ cov(z,x)&cov(z,y)&cov (z,z))

In Songhyun Academy the above case covariance matrix is used to assess the variables that increase student scores. For example, x is a number of hours a student spends reading, y is the number of times a student repeats reading a text and z is the height of a student. As the number of hours spent in reading increases, the performance score also improves and this is known as positive covariance. In positive covariance, both variables move towards the same direction. In negative covariance, both variables move in opposite direction then that is negative covariance and if a covariance is zero, then the variables are not related for example the height stated above is not related to academic scores hence covariance is zero.

References

Bamparopoulos, G. (2012) Statistical classification: A review of some techniques. Slide Share. https://www.slideshare.net/bamparopoulos/statistical-classification-a-review-on-some-techniques

Bhasin, H. (2016) . 4 types of Market segmentation and how to segment with them? Marketing 9. Retrieved from https://www.marketing91.com/4-types-market-segmentation-segment/

Biehl, M. (2009). Similarity-based clustering: Recent developments and biomedical applications. Berlin: Springer.