Saturday 27 September 2014

In a Excel file we are provided with 8 columns worth of data about soup ingredients. Columns A & B indicate the soup flavor where 1 = chicken...

Statistical analysis is the process of extracting meaningful information from raw data. There are two branches of statistics that are used in the development of the required information and these are;

  • Descriptive statistics

  • Inferential statistics

Descriptive statistics are used to provide a clear description of the data by obtaining information that would explain the data collected. They also show the distribution of the data as determined by descriptive statistical metrics. Descriptive statistics include; mean, standard deviation, coefficient of correlation, variance among others.


Inferential statistics are used to provide answers to hypotheses developed by the investigator or researcher and to offer predictive models. They also show relationships between the different variables based on the data collected. Inferential statistics include; regression, correlation, t-tests among others.


Starting with descriptive analysis, the mean for the different soup flavors and soup types can be calculated to determine the central value of the particular types and flavors.


The standard deviation is another measure that is calculated to determine the spread of the different values from the mean. A standard deviation close to zero shows the data is close to the mean, while those giving a higher value, show spread from the mean. This value can be calculated for the soup flavor and type independently by first calculating the variance then getting the square root where the result is the standard deviation.


The range is a statistic measure that is used to show statistical dispersion within the smallest interval in a data distribution. A range in this case can be calculated for the soup flavor and soup type.


Regression analysis which is a measure of inferential statistics can be calculated to estimate the relationship between the soup flavor and soup type. This can be done by holding the flavor or type as an independent or dependent variable. Regression would help the investigator to predict which soup type would be preferred for a particular flavor.


Correlation analysis can also be performed to determine the strength in the relationship between the different soup types and flavors.


T-test analysis can be performed to determine the difference between the soup flavors and types.


When developing the statistical analysis, it is advisable to start with frequency tables with percentages to provide a summary of the data and to enable simple extraction of the data for calculation using the models described above.

No comments:

Post a Comment

Is there any personification in "The Tell-Tale Heart"?

Personification is a literary device in which the author attributes human characteristics and features to inanimate objects, ideas, or anima...