Statistics can be defined as the practice or science that is concerned with the process of collecting and analyzing the numerical data in both large and small quantities. The process is enacted for the purpose of understanding proportions in complete from those in the representative data. Thus, there are several techniques that are used for visualizing relationships in relation to the data and also systematic techniques that give a fully understanding on the relationships that are concerned with mathematics. Therefore, in statistics there are several elements that are very important in regard to explain the entire concept and this include the following;
Descriptive statistics
Descriptive statistics are basically used to illustrate the sample that an individual is concerned with. They provide simple summaries that are relevant with the sample and measures. Based on the simple graphics analysis, they structure the basis of virtually each quantitative analysis of data.
Descriptive statistics is very significant since it helps to present quantitative descriptions in a simpler way and this is because in research study, there are numerous measures that are concerned with statistics. Therefore, one can measure a large number of individuals on any measure and descriptive statistics help to simplify those large amounts of data and making them more sensible in a way that they can be easily understood. Thus, every descriptive statistic lessens lots of data into a simpler summary (Winkler, 2009). For instance, if an individual has results of fifty pieces of student’s assignment, he will be concerned in looking at the general performance of those students. On the same note, he will also be interested in looking at the spread or distribution of the marks and therefore, descriptive statistics will help him to make the assumptions.
The aspect of properly illustrating the data based on statistics and graphs is very significant and therefore, there are two main types of statistic that help to describe data:
a) Measures of central tendency: this is basically concerned with describing of the central position in regard of the frequency distribution in a group of data. With that assumption, the frequency distribution defines the distribution and pattern of the marks scored by the fifty students from the lowest to the highest. Thus, one can easily describe the central position based on a number of statistics using the mode, median and mean.
b) Measures of spread: these are forms of summarizing a group of data by giving a description based on the distribution of the scores. For instance, the mean score of 50 students maybe 60 out of 100. But this assumption does not necessarily mean that all the students scored 60 marks and this is because some will be lower and others higher. Therefore, measures of spread help to make summaries on the spread of the marks and to describe the spread, there must be a number of statistics available which include the absolute deviation, variance, range, quartiles and also the standard deviation.
In addition, when using descriptive data, it is important to summarize groups of data using a combination of tabulated description, charts and graphs and also focus on the discussion of the results based on the data.
Inferential statistics
Inferential statistics is concerned with trying to reach effective conclusions that are extended beyond the immediate data alone. For instance, inferential statistics are used to help infer from sample data that gives an assumption on what the population have in mind. In addition, inferential statistics can be used to make precise judgments of the probability which an observed difference between different groups is dependable or on what might have happened during the study. Thus, the inferential statistics helps to make inferences based on the data or enhancing more general conditions (Asadoorian & Kantarelis, 2005).
Inferential statistics are very significant in experimental and quasi-experimental research design and evaluation. In fact, one of the simplest inferential tests can be useful when an individual wants to compare the average performance that is attributed from the two groups on a single measure to detect whether there is a difference. Most of the inferential statistics emerge from General Linear Model.
Hypothesis development and testing
Hypothesis is a statement that articulates concrete relationship among variables. It describes empirically testable connection among concepts and in many academic research projects, it is intended that the researcher should show competence in creating the foundation of the research project (McBurney & White, 2010). Thus, the research proposal always entails elements of the direction of the hypothesis.
Hypothesis is very significant as it gives the direction in a particular direction, it eliminates trial and error during the research and also helps rule out when confounding and intervening variable. Developing hypothesis needs divergent thinking in order to make sure that it is taken into consideration and this needs that redundant and irrational hypothesis is avoided in order to develop a good hypothesis. Thus, there are two significant qualities of hypothesis which test whether it is testable or falsifiable. Therefore, hypothesis development is basically based on the ultimate experience and this is attributed with reasoning, expression of both new and previous knowledge.
Based on the errors that are associated with probability sampling techniques, it is important to do hypothesis testing. This is because the researcher wants to measure whether the difference among the samples and the population properties are too low or too large. For instance, it shows whether the differences are vital to consider them more real compared to others than the sampling errors. Thus, the properties in this case include the mean, median, variance and mode (McBurney & White, 2010).
Selection of appropriate statistical tests
Selecting appropriate statistical test is very significant for the analysis of a particular research data. Therefore, using of inappropriate statistical test is commonly observed and can be seen in numerous conditions which involve using o paired test for the unpaired data. It can also result from using a parametric statistical test which does not necessarily follow the normal distribution of the statistical tests (Sharma & Petosa, 2014). Thus, due to the availability of numerous types of statistical software which help in the process, it becomes much easy however the statistical test still become the main challenge. Selection of appropriate statistical tests depends on whether the kind of data that is being incorporated, aim of the study and lastly, whether the data follows the distribution.
Evaluating statistical results
Statistical analysis is basically a quantitative method that is used to define the probabilities among the results of the data. Therefore, by evaluating the statistical results, the data can either be evaluated from either the social sciences or nature and the evaluation helps to analyze and elaborate the trends or the patterns within the research (Sharma & Petosa, 2014). Evaluation of the statistical will analyze the qualitative aspects that are meaningful in the research and all research summarizes the data by evaluating the data means assessed from different perspectives.
References
- Asadoorian, M. O., & Kantarelis, D. (2005). Essentials of inferential statistics. Lanham: University Press of America.
- McBurney, D., & White, T. L. (2010). Research methods. Belmont, CA: Wadsworth Cengage Learning.
- Sharma, M., & Petosa, R. L. (2014). Measurement and evaluation for health educators. Burlington, MA: Jones & Bartlett Learning.
- Winkler, O. W. (2009). Interpreting economic and social data: A foundation of descriptive statistics. Dordrecht: Springer.
