How to Use Microsoft Excel in Computing the Confidence Interval

Instat and GraphPad Prism are the ones that have the ability to compute for the confidence interval. This is part of most data analyses. However, the user needs to compute the intervals all by himself. In most cases, confidence interval is equal to the parameter minus or plus the error margin, which is definitely computed as the typical error times the exact critical value from its T distribution.  Additionally, the user can obtain the critical T value through the “free web Graphpad QuickCalc”. Keep in mind that if you just want to have 95% confidence, you have to enter 0.05 but if you wanted to have 99% confidence, then you have to enter 0.01. Through the use of Microsoft Excel, the user can compute certain confidence intervals. Thus, the critical T value for a specified number of degrees of freedom (df) as well as the confidence (C is usually 95%) is being computed through this formula “TINV(1 – 0.01*C,DF).

The mean confidence interval is totally centered within the mean sample then extends in the direction of the distance, which is equal to the critical T value times the mean standard error. SEM is equal to the sample SD, which is divided by the square root of the sample size. However, the user should be aware of Microsoft Excel’s “built-in confidence function”. It computes the confidence interval of the mean from another mean, which is the sample size and SD.

Prism, QuickCalc and InStat computes the interval from the SD sample. Through this way of computing, the confidence interval is much wider than the single computed by the confidence function in Microsoft Excel. Having large sets of data, the discrepancy is very little. Therefore, having small sets of data the discrepancy is large.


N and sample SD requires the use of the T distribution that the Prism uses. This is the way to compute using the confidence interval right from the mean sample. The N and population SD uses Z distribution or the Gaussian, thus, it computes the confidence interval which is from the mean sample. Keep in mind that the critical T value distribution depends on the sample size. Moreover, the critical Z value doesn’t depend on the sample size. Thus, the T distribution approximates the z distribution, when the sample size is considered large.

Confidence Intervals in Microsoft Excel are considered the interval which is around the sample mean in which the population’s mean to fall percentage of a certain time. The common levels of confidence function are 99%, 95% and 90%.

How to Use Confidence Function

The syntax of the Confidence function is “=CONFIDENCE(alpha,s,n)”. Alpha is considered the significance level. Thus, it is equal to one minus the level of confidence which is expressed as a decimal. The S argument is considered the standard deviation of the set of data. The N argument is giving the certain number of items within the sample.

Bear in mind that if the population value of the standard deviation is unknown, the user can use the sample value of the standard sample deviation, which is considered the standard deviation of the population.

Post comment