You are on page 1of 2

Statistical Analysis

Introduction: What is Statistical Analysis?

Statistical analysis is a collection method to process large amounts of data and result overall trends. S analysis is a useful tool to deal with noisy data. It provides ways to objectively report on how unusual is based on historical data. There are two different areas of statistics - descriptive statistics and infere statistics; which are related to but still different from each other.

Descriptive Statistics

tive statistics is simply the process of defining characteristics of a statistical measurement. Speaking, riptive statistics involves a observational study of a population. Charts and graphs are an important ro e standard measurements such as averages, percentiles, and measures of variation, and the standard xample in a paper reporting on a study involving human subjects, the table is giving the overall samp ple sizes in important subgroups, and demographic or clinical characteristics such as the average age, ortion of subjects with each gender, and the proportion of subjects with related co-morbidities.

Inferential Statistics

Inferential statistics is measuring the trustworthiness of conclusions about the population param on its information, this is called random sample. There are many possible uses of inferential stat example - political predictions. In order to predict who the winner of a presidential election is: ch sample from amount of Americans and asked which way they will be voting. From the answers g situation, statisticians will able to predict what general population will vote for with a high level o confidence. The keys of inferential statistics are choosing which members of the general populat polled and which questions will be asked.

Levels of Measurement
Once categorized according to the above scheme (i.e., independent, dependent, intervening), variables must be measured. This, of course, is the basis for data gathering. Data gathering employs measurement scales or sets of rules for quantifying and assigning values to a particular variable. Typically, four levels of measurement apply to data gathering. Data levels may be characterized as nominal, ordinal, interval, or ratio. The term nominal means to name. Hence, a nominal scale does not measure but rather names. A nominal variable thus consists of named categories. Examples of nominal variables include sex, religion, and "group" assignment (such as treatment/no treatment).

on a ratio scale. data should be gathered at the highest level. The reverse is not true. i. high level data easily can be converted to lower levels. Thus. . they also measure the distance between values. height and pulse rate. Common examples of ratio level measurements include a patient's age. Examples of variables measured on an ordinal scale would be pain levels (on a high-medium-low scale). Not only does the first patient have a higher temperature than the second. we can compare values not only according to the absolute interval between them. Whenever possible. Ratio measurement goes one step beyond interval scaling by providing an "absolute zero" point. Moreover. In other words. weight. For most statistical computations. or the rank ordering of patients according to their diastolic blood pressure. For instance. ratio and interval data are considered equivalent. but his temperature is 4 r C higher.e. The higher level of precision provided by interval and ratio data allows for more powerful statistical testing. ordinal or nominal. but also their relative magnitude. With ratio measures. an ordinal scale is a rank ordering of things.The term ordinal means to order. with a categorization in terms of more than or less than. Interval scales not only tell the order of things. assume you measure two patients' temperatures as 41r C and 37 r C. a variable with a value of 40 represents twice as much of the quantity being measured as a value of 20.