| Title | Short descriptive title. |
|---|---|
| Introduction | A paragraph or two to set the scene for the investigation. |
| Aim | One sentence or two to state what you are investigating. |
| Variables | Discussion of the key variables and their effects. |
| Prediction | Detailed quantitative scientific model (with diagrams) to explain what happens and why it happens. This may be referenced to secondary sources. |
| Method | Concise description of the procedure to be followed. |
| Diagram(s) | Scientific diagram of the experimental set-up. |
| List of Equipment | Specify type/size of equipment if this is important. |
| Risk Assessment | Hazards and the required precautions. |
| Variable Table | Summary of the independent and control variable values. |
| Trial data | Relevant trial data and an explanation of how it was used to decide upon variable values. |
| Appendix | All trial data, rough graphs and analysis etc. |
The summary above gives one approach to writing up an investigation, there are many possible variations. All investigations are different and each has a different emphasis of content. You must use your own judgement and common sense when deciding what to include.
There is no fixed format for writing up your Plan. As long as all the information is there then you will be awarded the appropriate mark. To achieve top marks, however, it would be expected that the write-up is in some sort of logical order and is easy to follow. Scientific words must be used, and used correctly. With confident use of correct science the write-up will not be very long, probably 2 or 3 sides of A4 paper. If it is any longer then you are probably waffling too much and don't really understand the investigation very well.
It is often a good idea to include all the rough trial data and analysis in an Appendix so that your teacher can double-check your work.
This is a guide to planning procedures to obtain the P.8 mark. A full understanding of the Higher Level knowledge from your syllabus is required. For an easier introduction have a look at the hints in the Mark Guide or read through the example investigation. It does not matter which science subject the investigation is based on, the approach is basically the same. Fieldwork is the only exception because you usually only get one chance at collecting data and samples.
Every time you start an investigation it seems like you haven't got a clue what to do. Don't worry, it's the same for everyone. Even your teachers would take a long time to carry out a new investigation such that they obtained full marks across the board, especially if they are not allowed to use anything beyond the scope of a GCSE syllabus. Often the best way to get up to speed is to try things out and see what happens - in essence you need to have a good play! It is surprising how much you can find out, and how quickly you can find it out, by doing this.
First you have to get a feel for what the investigation is about. At all times you are trying to find out what can be investigated and any of the key factors involved. You will probably just have covered a topic closely related to the theme of the investigation. Understanding the essential science behind the investigation is the key to obtaining high marks. Usually your teacher will introduce the investigation to the class via a group brainstorming session and most of the key ideas will be hinted at. It is a good idea to take notes on everything that is discussed. At this stage most people don't really understand the science behind the investigation very well, so it is important to have some notes to refer to during your initial planning stages.
You've probably got some idea about the investigation even though you don't fully understand it. Your next step is to try and sort out all the variables. You must make a list of all the possible variables. This will usually start as a small list and get bigger and bigger as you find more variables. Often something that you initially thought was one variable may be several closely related variables. Some variables have a big effect on the experiment, some a small effect, and some may have no effect at all.
For most experiments carried out in the classroom it is essential to carry out trial experiments. You can quickly carry out experiments to see what sort of effect the variables have. By quick, I mean quick! Don't waste a lot of time. It is much better to try things out than sit there thinking about them. Often you will find out your initial ideas are wrong anyway. Why spend 10 minutes wondering what will happen, when you can try it and get an answer in a couple of minutes or less? You will find that practical time is your most valuable resource. If you don't make the best use of it you won't be able to get top marks.
Get your ideas down on a bit of paper. The best way is to use a planning table. You should easily be able to fit all the main ideas for an investigation onto one side of A4 paper. If you've got big handwriting then learn to write a bit smaller when you need to. It is often an idea to do rough work in pencil so you can rub it out when you want to change things or to tidy it up a bit.
Start off with a simple table like this and make notes about all the variables.
| variable | values or range | how measured |
|---|---|---|
name or describe the variable |
Can the variable have different values or is it pretty much fixed for this experiment? If the variable can change what are the absolute maximum and minimum values for the variable? What are the practical maximum and minimum values you can handle with a typical experimental set-up? |
Explain how the variable can be measured. What equipment is needed? Is there a choice of equipment? What is the most appropriate piece of equipment, and why? What is the accuracy of the equipment? |
Example planning table entries.
A variable is anything you can change or measure in your experiment.
Some variables may turn out to have no effect at all on the outcome of the investigation. Most of these are usually quite trivial and can be ignored. Sometimes, however, it is important to establish that what appears to be an important variable does not in fact affect the investigation.
dependent |
The one you want to find out about. There may, in fact, be more than one dependent variable. Often they are all linked together and you can't change one without changing the other. The name comes from the fact that the value of the variable depends upon the values of the other variables. |
|
|---|---|---|
independent |
The one you change through a series of pre-set values. You have independent control over the values that you pre-set it to. |
|
control |
These must stay the same throughout the experiment to ensure a fair test. There are usually several control variables you must take into consideration. |
|
| Planning table example | ||
Each variable can be classified according to its type.
continuous |
Varies continuously and can have any value, limited only by the accuracy of your measurements. Examples: time; length; area; volume; mass; temperature; energy; force; |
|---|---|
discrete |
Changes by discrete amounts; i.e. in steps, or by one at a time. If the range of possible values is large then the variable may be considered as being almost continuous. Examples: number of layers; number of petals in a flower; number of paper clips; |
categoric |
Usually a descriptive category. Examples: colour; shape; size; type of material; |
For most Higher Level investigations there will be a choice of variables. If there is not suitable Higher Level knowledge to use for a prediction then choose another variable. You must be able to produce a good prediction to get 8 marks.
This is nearly always a continuous variable.
A Categoric variable is normally a poor choice. They are often the simplest variables to investigate but it is often impossible to generate a worthwhile prediction. In general, don't even consider a categoric variable.
A Discrete variable can get top marks, although you need to double check that what you think is a discrete variable is not in fact a continuous variable. Supposing you were carrying out an insulation experiment and decide upon the number of layers of insulation as the independent variable. You would not gain top marks. What you should do is measure the total thickness of the insulation. Thickness is a continuous variable and would give access to the top marks.
The best choice is often a Continuous variable. This allows you to put together a quantitative prediction.
If the investigation is too simple it will be impossible to gain top marks. If it is too complex you may run out of time or end up with poor quality results.
This is probably the most important consideration. If you do not budget carefully you will run out of time. Look carefully at what is involved to set-up and carry out the experiment, including all the repeats. Estimate the total time. Is it feasible? It is no good having a brilliant plan if you only manage to collect half the required results.
Check that all the equipment you need is available. Check that all the resources, such as chemicals, are available in the amounts you require.
The range of values you decide to investigate must be sufficient to give your prediction a good test. This means as big a range as is practical for you to carry out. You need at least 5 sets of data, 8 would be more realistic, more where results can be obtained quickly. In general, take as many measurements as you have time for. You need enough measurements such that there is no doubt that your graph shows a true and accurate trend.
Don't forget to check that you can still measure the outcome over all the values of the independent variable you intend to use. If the dependent variable goes off the scale you may be able to tweak some of the control variables to reduce it to more manageable values.
Readings must be repeated a suitable number of times. There is no magic rule for how many times you must repeat readings. If repeat readings are very similar twice would be enough. Most people take 3 sets to be on the safe side. If you get two similar and the third is way off you know something has gone wrong and you can double check it to eliminate the bad result. If there is a big variation between results then you must take more repeats to get a good average. You may find, for example, that repeats are close for small values, but get worse for large values. In this case use your common sense. Repeat say 3 times for small values, when the variation starts to get larger repeat 5 times.
You will already have carried out some quick tests to check out all the main variables. Now that you think you know which variables you are going to investigate you must collect some decent trial data if you do not already have it. Don't waste time taking repeat results if the data looks good. Experiment with values of control variables to get the best range of values for the dependent variable. Make sure that results are repeatable and not chaotic.
Once you have decided on suitable values for all the variables make sure you have enough trial data; e.g. one quick run through, maybe taking every other value, no repeats. For longer experiments, at least 3 sets (you will probably already have 1 set) - small, medium, big values, again no repeats. If there is a large variation between results then focus on a typical value and repeat it several times, 3, 5, 10 times, the bigger the variation the more times. This will give a good idea of the average value.
Make sure you take notes. You will need to select certain data to explain why your variables and range of study have been chosen. Tidy up all your rough notes so they can be included as an appendix of trial data in your final write up - they may prove crucial to getting a high mark. You don't need to spend a long time making them too neat.
There is no point taking trial data if you don't make use of it. You must analyse the data to see how good it is. So, draw graphs and see what they look like. See if the trial data fits in with your current science knowledge. Is the relationship between the variables as expected? Only by carrying out a full analysis of your trial results will you have any idea if your investigation will be any good. Usually a thorough analysis of your trial results will show where your method is weak and where you need to take additional results.
This may seem like a lot of work but remember that this is the most important part of the investigation. Getting to grips with the underlying science knowledge so that you can produce a quantitative prediction, and then designing an experiment to test this is crucial. Once you have done this, provided you have the patience and determination to obtain quality results, the rest should be easy.
A similar (if not identical) investigation to yours has been carried out many times before. It is far easier to sit down and learn from what has been done before than it is to reinvent the wheel. Start with your GCSE textbooks. They should contain the basic science theory if nothing else. Modern textbooks tend to have good pictures but can be light on information. Older books give more written information, simple A Level textbooks explain theory in more detail BUT may be difficult to understand. Books can be obtained from your science department, school library and public libraries, and there are cd-roms and the Internet. Your school and reference libraries also contain Science Data Books. These are usually big and heavy and have small print. There is probably one sitting on the shelf behind your teacher's desk. These reference books contain data for just about any science theory, equation, material or process you can think of.
When starting a search for information you will initially have to read everything. The more you learn the quicker you will be able to search. Learn to skim over what you already know in search of that essential piece of information. You can expect to find out many things that help you to understand but are not directly relevant to your investigation. Remember, when it comes to writing up your investigation, you must be selective with the information you include. Only include information that is directly relevant. With this in mind, don't forget to make a note of where you found that crucial piece of information so that you can go back and find it again.
The prediction is one of the keys to getting top marks. Science is not astrology so a prediction is not a guess or simple intuitive statement as to what you think will happen. A prediction is a scientific model, including Higher Level science knowledge, that describes what should happen and why it should happen. In most cases the prediction will also be quantitative. You should know what will happen because of the trial experiments you have carried out. In general you don't gain any credit for just stating what will happen.
Let us start with an example of a qualitative prediction;
"The dependent variable(y) will increase as the independent variable(x) is increased, because..."
This doesn't tell us about the increase. It could be a small or large increase, and the increase could be linear (constant) or non-linear (increasing by a smaller or bigger amount).
Contrast this with the start of a quantitative prediction;
"The dependent variable(y) is directly proportional to the independent variable(x) squared because..."
This tells us that the relationship has the form y=mx2, where m is a constant. The relationship has now been specified in numerical terms. The constant 'm' is not known and you will not always be able to predict it, although you will be able to work it out from your results.
You must be able to justify a prediction to gain top marks. By this it is meant that you can explain, using science knowledge, your quantitative prediction. To do this you will have to put together ideas from equations or laws. This is the hardest part of the investigation and will require quite a bit of thinking about. Don't expect to sit down and work out the answer in five minutes. First you must understand all the related science, then you must pick out the relevant bits, combine them together, and then apply the knowledge to your individual situation. Thus it could take you several days or a few weeks to put together a good prediction. The important thing to do is to revise it and improve it as you carry out your investigation and come to a better understanding of what it is all about.
Include diagrams to make your prediction come alive. A picture can often show clearly something that is difficult to explain with just words.
Remember that your teacher has carefully selected your investigation so that you are able to produce a prediction. If you think it is impossible, then ask them if this is one of the very rare investigations where it is not possible to produce a prediction, and can you still get top marks?
When you write up your prediction it is a good idea to include a sketch graph to show in a visual way your predicted relationship. A 'sketch graph' doesn't mean you sketch it freehand! It means you draw it neatly to show the shape of the graph and any important features such as intercepts. Sketch graphs are not normally done on graph paper.
Some predictions lend themselves to being modelled mathematically by equations. If maths is one of your strong points then you may be able to build up a good model to describe the relationship between your variables. If this is done with a computer package you may be able to work out what will happen for various alternatives, and then be able to print out a series of graphs. These models can be used to compare against the real data you obtain.
When you have used secondary sources of information it is important to acknowledge from where you got the information. You must do this for each source of information. Usually you can use abbreviations when referring to books (e.g. title or author) provided that you give full details elsewhere. If you have several sources it is best to collect them into a section titled "References" at the end of the write-up. If you only have one or two it is easiest to give the details when you use them.
In most cases you want to try and measure as accurately as you can. There are several factors that affect the accuracy of measurements and it is a good idea to understand what some of these are.
We can imagine when a measurement is taken that if we had perfect measuring equipment then we could obtain a perfectly accurate result. Suppose we have a piece of electronic equipment with "2.3" on the digital readout. Does this mean the value is exactly 2.3? Well, it could in fact be 2.34 or 2.336 and so on. The number of decimal places given by the display is one indicator of the accuracy of the measurement. In general, the more digits there are the better the accuracy is.
This is not the end of the story because no instrument can be 100% accurate. If we used five different instruments of the same type they may read 2.3, 2.3, 2.4, 2.3 and 2.4. Which is the best value, 2.3 or 2.4? It is impossible to tell, unless we had another more accurate instrument to check the value against. The manufacturer of the instruments would probably do this to check that they are all working correctly when they are first made. Provided that the instrument is correct within a certain manufacturing tolerance (e.g. 5%) then it will be packed up and sold as a good instrument. If the instruction booklet has not been lost then you can usually check what the measurement accuracy is. It is often given as a percentage. The more money you spend on an instrument usually the better the accuracy is. This does not just apply to electronic equipment but anything that is used for measuring. If you check several cheap plastic rulers, for example, you will often find that they are different over 30cm by more than one mm.
Another factor that limits the accuracy of measurements is the person taking the measurements. Some equipment must be calibrated before it can be used. If this is not done carefully then all the measurements recorded may be significantly wrong. It goes without saying that equipment must be used correctly to get the best results. A high quality instrument may produce rubbish for a sloppy or careless operator.
Reliability is crucial in any investigation. If at the end of the day you can't reach a reliable conclusion then your time will have been wasted. An investigation is reliable if anybody can repeat the same investigation and come to the same conclusions, within the same accuracy. The key to producing a reliable investigation is in identifying all the variables and then choosing appropriate values for the independent and control variables. Not only do you need a "fair test" but a suitable choice of equipment, a reliable method and good practical skills.
A method must be chosen to produce reliable and accurate results. Judgement is required to select equipment and set it up correctly. The best method is usually found by trial and error, and by improvisation of standard practical techniques. Good practical technique means setting up equipment so that it is safe and tidy and doesn't require you to be double jointed or have three hands to take the measurements! Allocating specific jobs to people and working well as a group also helps.
When writing a method keep it simple and straightforward. Steps should be presented in a logical order, with nothing missed out. Always explain any techniques used to increase the accuracy of measurements, or to make measuring and recording easier. Refer to diagrams when this makes the explanation clearer.
Every experiment must have a risk assessment carried out for it. Consider everything about the experiment and decide if there are any safety risks. If there are any risks then you must be able to control them to minimise the chances of an accident.
Any time you are using chemicals you need to check the possible hazards. Your school may have Hazcards which list all safety issues and the precautions needed. Consult these if they are available, otherwise you will need to ask for specific safety information. In general, safety specs or goggles are always required when using chemicals. Other precautions could be a safety screen, working in a fume cupboard, wearing disposable gloves, or using small amounts.
Other common hazards are electrical short-circuits, burns from hot objects, sharp objects, unstable equipment, falling objects, objects under tension suddenly snapping, and harmful microbes.
Every experiment is different and it is not usually difficult to spot what the risks are and the precautions that need to be taken. Remember that in the Obtaining Evidence skill you are judged on whether you actually carry out the investigation safely, not just whether you included the safety at the planning stage!
When writing up the risks you do not need to include trivial safety issues such as how to light a Bunsen, remembering to tie your hair back, or being sensible and not running around the lab. Similarly, if there are no safety issues then write just that.
A labelled diagram must go with your method. Often it is easier to draw this by hand (neatly using a pencil and ruler) than it is to try and draw it using a computer package. In general these should be technical science drawings using standard symbols for common equipment such as beakers and flasks. You should have lost the habit of sketching stopwatches and Bunsen burners well before Year 10! You can sketch things where more detail is required to show exactly what is happening, particularly for non-standard equipment or set-ups.
Make a list of all the equipment that is needed, giving specific details of the size or type of equipment where this is necessary.
To finish off the method it is a good idea to summarise the values to be used for all the variables. This is easiest to do as a small table, similar to the planning table. For the independent variable you must show the range and number of measurements, and how many repeats are to be taken. The single value for each control variable must also be listed.
If you have used trial data to help decide upon the values to be used then you must summarise the relevant data and explain how this helped you to decide. To gain high marks you must clearly show how the trial data has helped you to choose the best possible range for the independent and dependent variables. The best values are those that will enable you to thoroughly and reliably test your prediction.
Planning is pretty much the only skill that can be carried out completely on its own. There are some advantages to this in that you can plan an investigation without worrying about having all the necessary equipment and the time to carry it out. The disadvantage is that without trying out the experiment you may not understand it so well and may end up making a silly mistake in the plan.
As you don't have to worry about time always plan to take enough readings to show the best possible relationship. Plan to be flexible in your approach; e.g. "Taking readings every minute should be sufficient, but if there are any sudden changes readings can be taken at much shorter intervals to more accurately follow the change." You can also plan to modify control variables to find out the best values to use, provided you explain why you need to do this.
To finish off your plan it may be a good idea to draw up a blank results table showing exactly what you plan to measure, including the correct units. Don't forget to list the value of all control variables if you have not already done this.