How can we show a 95% confidence interval on a graph? The first step is to make a bar graph. In this exercise, we will graph the sample means of the different-sized samples you collected in the last part of the activity. Use the grid below. (Note - it is important that bars are two or three squares wide and that you DO NOT shade in bars) Now it is time to illustrate the range of means within which we are 95% confident the true mean should be found. This is done by including what are called 95% confidence interval (CI) error bars to the bars representing your sample means. One error bar should extend 2 SE- above the sample mean and the other error bar should extend 2 SE-below the sample mean. See the example to the right. Sample mean: 57.8 kg Standard Error of the Mean: 4.7 kg 95% Confidence Interval: 9.4 (2 SE-) 26. Describe the amount of overlap you see in the 95% confidence interval error bars from the different sample sizes. Summary: When 95% CI error bars from two samples overlap, we cannot claim that they have different means. This would make sense in this case because all of the samples were taken from the same population. When the error bars between two samples do not overlap, we can say that the means of the samples appear to be different from each other. Body Mass (kg) 80 70 60 50 R2222 30 20 10 4