Lesson 20 Treatments and effects
Teacher guide
THINKING SCIENCE Teacher’s Guide
Treatments and effects
OVERVIEW
In physical sciences relationships between variables can often be described unambiguously (‘proportional to’, ‘inversely proportional’, or defined in terms of an equation). As the woodlouse activity (Lesson 19) emphasised, living things incorporate so many uncontrollable variables that their behaviour cannot be reliably predicted. Thus to explore possible relationships between causes and supposed effects is not straightforward. Correlation reasoning is the process used to assess the strength of relationships between variables, and this activity provides examples to show pupils what is meant by correlation. An understanding of correlation relies on an understanding of probability. The reasoning is like this: ‘if the effect is not due to chance then what is it due to? We look for the strength of a link (a correlation) between cause and effect because some of the effect may be due to chance’. The results of experiments carried out on various plants and animals are used by students to construct a 2 ⋅ 2 contingency table, to look for positive, negative, or zero correlation between treatments and their effects.
Most of this lesson is done as a whole class with short amounts of time given for paired discussion. The last part of the lesson allows for consolidation and construction when pupils use cards to look for correlation relationships.
New key words: correlation, prove, disprove
EQUIPMENT REQUIRED
Per group
Notesheets.
Pack of 20 treatment/effect cards. (Material for five different packs of cards is provided on the Workcards.)
Templates 20A, 20B, 20C, 20D, 20E
LESSON PLAN
1 Introduce an example. ‘We want to see if spraying carrot plants with a new chemical called “Grocaro” makes them bigger. If it doesn’t affect them, or if it makes them smaller, we will stop making the chemical and try to find a new one instead.’ Show the overhead picture of carrots but keep the results box covered at this stage. Ask these questions:
What are the input and what the outcome variables? Does ‘Grocaro’ make carrots bigger? Count them up. What do you think?
‘Grocaro’ is the input variable, or treatment; size of carrots is the outcome variable or effect.
What do they think about the effect of ‘Grocaro’?
Let them discuss in pairs for 2 minutes. Now take votes. Most will vote for it working, ask three or four pupils to say why they think that. This is important for comparison and consolidation before the cognitive conflict. (5 minutes)
2 Now introduce the cognitive conflict. Ask ‘if I use “Grocaro” in my garden, will all my carrots grow bigger?’
Some pupils will say yes, some say no, most say some carrots will. Challenge them, ‘but if it works I would expect it to make all my carrots grow’. Are you saying it sometimes does not work? How do I know if it is worth using?
Maybe my carrots would be that big anyway?
Some pupils will point out where the carrots don’t fit with expected results; i.e. small carrots with treatment and large carrots without treatment.
Now introduce the results table and summarise the numbers of each category in the results boxes on the overhead.
This does seem to show that ‘Grocaro’ improves the size of carrots. There is no need to labour this point now. It generally provides some

Treatment with ‘Grocaro’ 
No treatment with ‘Grocaro’ 
Effect: bigger carrots No effect: usual size carrots 
11 
7 
cognitive conflict to have to count untreated plants as well as treated ones, and pupils will get practice applying the principle to their own results. You should, however, introduce the word correlation as a description of the relationship, with a positive correlation meaning that the treatment does have an effect. (5 minutes)
3 Show (OHP transparency 2) Now look at ‘Gleamo’, the amazing shoe polish.
Ask what results you would expect if ‘Gleamo’ works? Let pupils discuss this in pairs for 1 minute then get some ideas from them.
Systematically count all ‘“Gleamo”, yes’ boxes (treatment gives effect) and all “No ‘Gleamo’, no” boxes (no treatment, no effect). Pupils can do this for you with coloured pens on the OHT.
What does this show?
At this stage some pupils should suggest looking for disproving cases; i.e. all ‘“Gleamo” and no effect’ results and the ‘no “Gleamo” but shiny shoes’ results.
Complete the results table. What do they conclude now? This is a negative correlation. Again there is plenty of opportunity for cognitive conflict and construction. (10 minutes)
4 Put the pupils into groups of four. They are going to be scientists analysing the results of some important experiments on animals and plants to see what treatments will increase the productivity of various plants and animals. Give each group a set of cards and a Notesheet. Define each ‘investigation’.
Sheep having red pills to try to increase wool yield.
Cows getting green pills to increase milk production (in bottles).
ANSWERS
Results for each set of cards Sample size for each set is 20 organisms.

+T +E 
T +E 
T E 
+T E 
A+C: B+D 

A 
B 
C 
D 

Sheep/ red pills 
6 
4 
6 
4 
12:8 
Cows/green pills 
4 
6 
4 
6 
8:12 
Pigs/brown pills 
7 
7 
3 
3 
10:10 
Wheat/fertiliser 
5 
4 
6 
5 
11:9 
Roses/Leaf shine 
3 
3 
7 
7 
10:10 
Pigs getting brown pills to make them fatter, so more meat.
Roses having their leaves sprayed with leaf shine to make more flowers grow.
Wheat having fertiliser added to increase the amount of grain (in sacks).
Explain how to sort the cards into four categories, counting and filling in the table with the total for each category. Each group should aim to do at least two packs of cards; i.e. ‘analyse’ the results of at least two investigations.
Once they have their results in the 2 ⋅ 2 tables the students should find out if the treatments seem to work or not.
Looking at any two cells with a complementary attribute (e.g. plus treatment, plus effect and no treatment, no effect), can they say ‘the treatment had an effect because more plants/ animals which had the treatment show an effect than plants/animals that did not have the treatment’? If they cannot give a very positive answer to ‘do you think the treatment works?’
(and sometimes the figures ensure that they cannot) they must go on to compare two pairs of cells to make two comparisons. One point of debate is ‘do you have to have the same number of organisms in both the control and the treatment groups?’ The answer is, not necessarily.
Comparing ‘proving’ cases with ‘disproving’ cases shows the type of correlation to be found. Emphasise the word correlation. Keep asking if the effect could just be chance or is it the treatment? (20 minutes)
5 Finally, ask different groups to summarise the different investigations. Give practice in the use of the terms ‘positive correlation’, ‘no correlation’. Bridge to simple medical examples such as the possible effect of aspirin on headaches, or of cream in reducing spots.
(10 minutes)