**Abstract**

Confidence intervals are one of the statistical tools for conducting hypothesis testing. Hypothesis testing is a common rule in research for testing the validity of the unknown population parameter. There are many different types of test-statistics used in hypothesis testing. Some are more appropriate and specific to some problems depending on the variables being compared. For instance, Chi-square works best in testing relationships of categorical variables. However, whatever appropriate method employed in the research hypothesis, the Null hypothesis is a reliable concept for testing an assumption. The null hypothesis (H0) assumes no difference or no change in variables (Sullivan, 2012). In contrast, the research question or hypothesis (HA) is a set of educated guess the researcher believes to be true within the context of the research (Sullivan, 2012). This review explores the application of confidence intervals in hypothesis testing.

**Sampling and Confidence Intervals**

The confidence level shows how sure and reliable the data is. It is expressed as a percentage and represents how often the true percentage of the population. The higher the confidence level is, the more certain or confidence attributable to an estimated value or data. As the confidence level increases, the lower the margin of error a researcher is willing to accept (Usable Stats, 2014).

The confidence interval is also called the margin of error associated with the plus or minus sign with the calculated mean value or research-question’s value. The wider the confidence interval is, or one is willing to accept, the certainty or reliability of the whole population falls within that range (The Survey System, 2012). More importantly, confidence interval is affected by three major factors; sample size, percentage, and population size. The larger the sample size, the more reliable the confidence interval truly reflect the population. Consequently, for a given confidence level, the larger the sample size, the smaller the confidence interval, however, the relationship is a non-linear relationship (The Survey System, 2012).

Perhaps, having established the basis and critical preference criteria on sampling, confidence level and confidence intervals, the concern is in maintaining the integrity, validity and reliability of the study’s assessment whenever the confidence interval or confidence level changes. Therefore, imagine that investigators wanted to suggest or predict that the percentage of Americans who prefer bottled-water fall between 62% to 67% confidence interval using confidence level of 90% as opposed to the 95% confidence level for the 60% to 70% confidence interval. Clearly by using 90% confidence level with 62% to 67% confidence intervals, it reflects a narrow interval difference of 5 instead of 10 observed with 95% confidence level with 60% to 70% confidence interval. Therefore, confidence intervals are critically analytical tools in the health science and public health field employed to validate the reliability of data or statistical analyses in research studies.

References

Sullivan, L. M., (2012). *Essentials of Biostatistics in Public Health* (2^{nd} Ed.) Burlington, MA

Jones & Bartlett Learning pp. 101-118.

The Survey System. (2012). Sample size calculator. Retrieved from http://www.surveysystem .com/sscalc.htm

Usable Stats. (2014). Introduction to confidence intervals: Relationship between the confidence Level & confidence interval. Retrieved from http://www.usablestats.com/lessons/ Confidence LevelMore

Walden University. (2014). P-values and confidence intervals. Retrieved from https://class. waldenu.edu/webapps/portal/frameset.jsp?tab_tab_group_id=_2_1&url=%2Fwebapps%2Fblackboard%2Fexecute%2Flauncher%3Ftype%3DCourse%26id%3D_4757972_1%26url%3D

Comments are closed.