How to plan an experiment

​​In a nutshell

Planning an experiment is one of the most important steps when it comes to physics. Every experiment requires different apparatus, are measuring different quantities and will test you on different techniques that you need to know. When testing a particular theory, you must research it and plan so that you know exactly what sort of experiment you will be conducting.

Definitions

​

Using appropriate equipment and techniques

When planning an experiment, research is done initially to figure out which apparatus is appropriate for the experiment in question. A light gate, for instance, is a timing mechanism that is used in experiments that require timing and speed calculations of objects, and you would not use it when investigating a gas.

Knowing the independent and dependent variables of your experiment is important. The independent variable is the quantity that will be changing, as it doesn't depend on anything else. The dependent variable is the quantity you will be measuring, that changes as a result of the independent variable changing.

Example

What are the independent and dependent variables when measuring how the angle of a ramp changes the speed a trolley reaches the bottom at?

The quantity being measured is the trolley's final speed. It is dependent on the angle of the ramp, so it is the dependent variable.

Meanwhile, the angle of the ramp is being changed. This is the independent variable.

In addition, you must make a judgement on the scale of the experiment. In an experiment investigating the speed of a trolley that is $$15\,cm$$ long, you would not need a ramp several meters long. 

Similarly, using an incredibly small pendulum with a very fast rate of oscillation would make taking measurements very difficult.

Example

For measuring the instantaneous velocity of a trolley on a ramp, what equipment would you need?

First, you need the trolley and ramp themselves.

As the experiment is searching for the velocity of the trolley, you need a way to measure time and distance, so speed can be worked out.

For time, you would use a light gate connected to a data logger, and for distance, a simple meter ruler will be fine.

Identifying variables

Identifying other variables involves figuring out what quantities could change and making the effort to control them. The research done on constructing experiments also helps you to identify other variables in the environment.

If a variable isn't controlled, then the results of an experiment will be hard to decipher. Changing values could be attributed to either the purposefully changed variable, or the uncontrolled one, and it would be hard to tell which.

Example

What are some variables that must be kept controlled when investigating acceleration of a trolley down a ramp?

The acceleration of the trolley would be affected by the coefficient of friction of the ramp, as well as the angle at which the ramp is situated. Therefore, these need to be controlled to get reliable results. 

Evaluating the experimental method

Evaluating the method of the experiment involves making sure that the method that was used is a valid way of doing the experiment. One of the most important ways this is done is by repeating the experiment. Checking that the experiment produces similar results each time you repeat it is a good way to tell if it is valid or not.

Experiments should be constructed so that they produce as high accuracy and precision as they can, while removing limitations. An experiment with high accuracy describes an experiment that produces readings and measurements that are close to its true, physical value, while precision is used to describe how far or close repeat readings are to each other.

It is important to check the experimental process for any of these limitations, and make sure they are removed or corrected.

A All the points are near the bullseye, making them accurate. They are also near each other, making them precise. B The points are near each other, so they are precise, but are nowhere near the bullseye, so they are not accurate.

The accuracy of an experiment can be affected by systematic errors - for example, a scale that has not been zeroed would be a source of systematic error. 

Accuracy can be improved by using instruments with an appropriate measuring scale. Each instrument has a particular resolution, which describes the smallest variation it can measure, and using an instrument with the appropriate resolution will increase accuracy.

The precision of an experiment is affected by random errors - using the equipment differently for each repeat of the experiment may result in lower precision.

Example

When taking repeat readings for a trolley on a ramp, what are some mistakes that you could make that would lead to decreased precision and accuracy?

Placing the trolley in different positions on the ramp at each time. Changing the trolley's initial position will affect the final speed, and thus will decrease precision.

Using a short ruler for a meter-long ramp would lead to decreased accuracy, as the ruler's resolution is not appropriate for the length of the ramp.