Lab Tips and Skills
An example Lab Report
Not all labs are the same, but here is an example lab I put together for students that were asked to design a lab about whether or not the projectile motion model accurately predicted projectile range.
Identifying a Problem
Every lab should be investigating something that should be written in a question form that can be answered through a controlled experiment. Some labs are verification labs (where we test a known relationship) while others are exploration labs (where we try to identify a relationship we did not already know), but the difference may not be apparent in the problem.
- What factors affect the rate of free-fall motion? (probably exploratory)
- What is the acceleration due to gravity on earth? (verification -- it should be 9.8 m/s2)
- Is energy conserved in an elastic collision?
There are 2 ways to write procedures: a formal procedure set and an outline. We will most often write outlines in the interest of time so we can provide more lab experiences.
All raw data is recorded in a table format. It is OK to use one format for data collection and include a place for averages there, then reformat the table to make analysis easier later. You should include both if you do that.
While recording your independent variable and dependent variable data in a table is essential, you should keep your eye out for any other observations that might affect your results or even just be interesting.
- You can notate data points with an asterisk and make a note at the bottom of the data table
- You can write a brief paragraph or include diagrams to indicate other observations.
examples to come
Choosing the right graph
- Always include a literal calculation that shows what the slope and/or area of the graph represent. If multiple values are referred to, create a new data table to summarize them.
- For repeated calculations, show 1 example in long form (equation, substitution, answer with units) and then a data table summarizing results if you wish. Averages do not need work as long as they are in a logical place.
- Show work for any values you will refer to in your analysis summary/conclusion.
- Show % error or % difference calculations if used.
- Conclusion should not introduce new data/calculations, but should refer to them
There are 2 types of error we often run into.
1) random error - average of measurements is close to true value (averaging multiple trials reduces error)
ex. most stopwatch error could be either anticipating or delayed reaction. Averaging them is a good way to reduce error.
2) systematic error - true value is higher or lower than the average measurement (multiple trials do not help reduce)
ex. speed of a car that curves between measured points - average of multiple trials is always lower speed than true speed because the actual distance traveled is further than the recorded value. Averaging will not reduce this error. We can either find a way to include the actual path or logically justify the reduced speed.
When you have random error, you report the average, but should really understand that any number that falls within the range of that data set would be considered to be statistically equal.
A motion is measured with the following times recorded for the same motion
The value to be reported would be 6.5 s ±0.3 s because the range of accepted measurements includes values up to 0.3 s off from the average.
You should not be required to report your measurements this way on all labs, but you should understand it and may choose to use it as a statistical tool when analyzing your own data.
Quantifying the quality of your results:
Rather than use vague statements, like "pretty close" or "very accurate" you can find % difference or % error to give a numerical value to your statements. This is a much stronger way of supporting your claims than an opinion.
- if 2 values are supposed to be the same or you are comparing them, use a % difference
- if you are verifying a value (testing against a predicted value) use a % error