Statistical Significance

 Introduction

            In simple terms statistical significance denotes the likelihood that the results of any finding are caused by another factor rather than chance or a chance influenced factor. This term was first used by Ronald Fisher, and it does not denote the literal sense of meaningfulness or importance. The significance level is commonly set at level less than five percent (5%) with a probability of less than 0.05 (p<0.05). This basically implies that the findings’ results are likely to be 95% accurate or the results may have been caused by a chance of at least 5% percent.

Statistical significance tests are used to address issues of doubt about results from analysis of postulated relationships and associations. There are two important questions that are always asked about assessed and analyzed relations: 1. what is the level of probability that shows that indeed the relation is existent? 2. What is the strength of a relationship, if at all it exists within a postulated hypothesis? These tests of significance particularly address the first question. They show us the level of probability that the relationship we have established is not just a chance occurrence. They also show us the level of chance that implies that we would be erring to state that the relation we have established is indeed true. They also show us the level of confidence with which we can state that the relationship applies to other randomly sampled samples as well as populations (McCloskey & Ziliak, 2008).

There are different types of statistical significance testing modes. These include the chi square test. This type of statistical significance testing is used in the analysis of ordinal and nominal types of data. Chi square tests are termed as non parametric tests which compare findings from research to the anticipated results hypothesized (‘PPA 696 Research Methods’, 2010). It is important to note that this test does not show level of association and association direction between compared variables. The test does not also show type one error probability, a concluded and absolute proof about the relationship. Neither does the test account for research validity and reliability (Carver, 1978).  Another commonly applied test of significance is the T-test. The t-test significance test is applied in the analysis of ration level and interval data. This type of test is applicable to varying kinds of statistical tests.

The first use is in the testing of differences that exist between two samples considering only one variable. This test is based on the value of the mean of the variable under assessment for the two samples. An example would be to compare scores from students in the same grade but in different schools (e.g. public and private schools). The t-test can also be used in establishing whether the mean value of a sample is higher or lower compared to some standard value. An example would be testing whether the average score of students in private schools is higher than 50% percent. The t-test is also utilized in testing whether one sample has varying means in different variables within the same sample (‘PPA 696 Research Methods’, 2010). An example would be to establish whether the same sample of clerks is more efficient and fast on Ubuntu or Linux operated systems.

These tests are applied in virtually all forms of research and statistical analysis in order to establish validity of research. The tests of significance are particularly important in validating data and hypothesis postulated within any form of research. Conclusively, any research conducted without a final analysis on the significance of the results may be declared null and void.

References

Carver, P. R. (1978). The Case against Statistical Significance Testing. Harvard educational Review Journal, volume 48, issue number 3, pp378-379.

McCloskey, N. D. and Ziliak, T. S. (2008). The cult of statistical significance: how the standard error costs us jobs, justice, and lives. University of Michigan Press.

‘PPA 696 Research Methods’, (2010). Tests for Significance. Retrieved on 19^th August, 2010 from http://www.csulb.edu/~msaintg/ppa696/696stsig.htm.

Is this your assignment or some part of it?

We can do it for you! Click to Order!

Is this your assignment or some part of it?
We can do it for you!