In an experiment (gravimetric analysis of sulfate), I had to find a sulfate composition in an unknown salt by
mixing it with BaCl2 in a solution. The answer was 60.33% or 59.4 +or- 1.00% ( confidence interval). My results were absurd compared to the answer (54.77 +or- 21.44%). There were three trials and the results were (57.90, 61.40, 45.01). Obviously the last result caused this huge amount of confidence interval, but it did pass the q test. The last result was caused because I accidentally knocked over the crucible and spilled some of the precipitates, but I did get some of the precipitate back in the crucible. My questions are, how should I describe the quality of the experiment and the measurements, did I get the right confidence interval and if the confidence interval is large, how should I justify this in a report.