Sample Lab Report
Writing up a Lab Session
The sessions in Doing Data Analysis with Minitab 12 are designed primarily to help you understand and think about the statistical techniques we are learning. Minitab is a tool which can expedite analysis, and give you more opportunity to focus on statistical thinking. As such, the general goal in the sessions is to find the "story in the data." As you follow each set of instructions, pay attention both to the Minitab procedures and to the meaning of the analysis.For each lab, you will prepare a brief "Lab Report" including these elements:
Answers to the italicized questions posed in the session
Answers to the assigned "Moving on…" questions
Computer output related to the assigned "Moving on…" questions
Any comments or questions you have about the software, the statistical techniques, or interpretations, as needed.
So that the lab sessions are not merely mechanical activities, each one contains questions for you to answer. Some of the questions are rhetorical or intended to lead you through a thought process. There is no need to write answers to these questions. However, you must answer questions which appear in bold, italic type within the lab manual, as well as any "Moving On…" questions that you were assigned.
As you go through each session, you will also generate other output which should help you to become familiar with Minitab and with the statistical concepts. You should include some of the output in your report, but only attach those pages of output which helped you decide on your answers for the Moving On questions.
Attached is a sample of a typical lab report which would be appropriate for Lab Session #2, prepared by two students who were assigned Moving On… questions 2 – 4. Note that they start by discussing a point of confusion; this is quite appropriate, but obviously not required in every lab report. Note as well that the students indicates how they approached the Moving On questions, by citing the technique used to arrive at an answer.
Also: to save paper, I have "cut and pasted" Minitab output into a word-processed document; ordinarily, you would simply staple the original output to your lab report. Reports must be grammatically and statistically correct, and should be "businesslike" in appearance.
Sample Report
- B. MacPhaidin
- R. Kruse
- BA205 Quant. Analysis I
- Prof. Carver
- September xx, 19xx
Lab Session 2
This lab concentrated on histograms and other charts to summarize data. The only problems we ran into came up when we had to decide how many intervals to include in the histogram for insurance premiums. We weren’t sure if we did this right, but after talking to some other people in the class, decided to choose the one shown below.Session Questions
(14) The distribution is strongly skewed to the right. The outlying dot represents a country with considerably more AIDS cases (about 13,000) in 1994 than any of the other countries. We think it may be the United States, and we looked in the data window to verify that the USA had 13,355 cases in 1994.
(16) The histogram is very similar to the dotplot; it is harder to notice the outlier in the histogram than in the dotplot.
(17) The second histogram has the same horizontal and vertical scales as the first, and the general shape is very similar. The main difference is that the second, with more bars, shows more detail in the variability of Case94. It’s hard to say which summarizes the data better, but I’d choose the second, since the first one "smoothes over" the variation for countries with under 5,000 cases.
(18) In the Rate93 histogram, there is a positively skewed distribution again. It looks like most countries had rates of under 20 cases per 100,000 population. Once again, there is an outlier, but it is Zambia this time (not the USA) which has the highest prevalence.
As the name suggests, this graph shows the Cumulative Frequency of different rates of AIDS cases. That is, for each value on the horizontal axis, the height of the bar shows the number of countries with that many cases or fewer. Reading off the graph, it looks as though about 200 countries have less than 50 cases per 100,000 population.
(20) According to the tally, the most common means of transportation was car, truck, or van. There was a tie for least common, with subway and bicycle each having one ‘vote.’ Of course, in this sample, no one responded 3 (Streetcar or trolley), 6 (ferryboat), or 8 (motorcycle), making those even less common.
(21) The chart and the tally do contain the same information. Each bar in the chart is proportional in height to the "Count" column of the Tally.
Moving On Questions (2-4)
States.MTWWe used a tally command, displaying counts and percents, shown below. We conclude that 23.53 percent of the states (i.e., 12 states) set the level at .08.
We chose to use a histogram here. The histogram of premiums shows that average premiums varied substantially across the US, from roughly $400 to $1100, with typical premiums running in the range of $550 - $750.
From the cumulative frequency histogram, it looks like only 3 or 4 states had more than 1500 fatal accidents that year.
Sample Output for the Moving On… questions:
Summary Statistics for Discrete Variables
