Anna asked:

I’m tutoring a practical class for first year biology (uni), and I have taken issue with something they are teaching the class regarding hypothesis testing, I would like a second opinion.


In their very first practical my students were given a hypothesis:

‘if red blood cells contain haemoglobin (which is purple), then the organ that red blood cells are stored in will be purple’

(Obviously this assumes that purple haemoglobin would turn an organ purple, but let’s just take this as a given).

After being given this hypothesis, they were then instructed to investigate by dissecting a toad and trying to figure out which organ contains the red blood cells from visual inspection, i.e look for the purple organ. The problem is (the root of the entire mess, which I will describe shortly) that a number of organs are purple, not just the organ that contains the red blood cells (the spleen), therefore it’s not immediately apparent from visual inspection alone which organ contain the red blood cells, unless you happen to know beforehand that the organ is the spleen and so can just look for the spleen and confirm that it is indeed purple (but they aren’t supposed to know that in advance). So due to this not being apparent, the head tutors are claiming that this hypothesis is not falsifiable.

I do know that something has gone wrong here (general flaw of induction perhaps?), I suspect that it’s something to do with the relationship between the aims (locating organ that stores red blood cells) of the study and the hypothesis. If someone can clarify where exactly things went wrong and how, thanks, my brain is just not getting there. However I don’t see it as being a problem of ‘lack of falsifiability’. Claiming that this statement isn’t falsifiable is (IMO) like saying ‘if coke is black, then a clear glass with coke in it will look black’ is not a falsifiable statement just because other black liquids in clear glasses will also make those glasses look black.

Am I right in thinking this or am I missing something?

Can I get some insight from some epistemologically savvy people think before I put my reputation on the line with the course co-ordinator? Apologies if I’ve totally muffed this up.

Answer by Shaun Williamson

Anna you need to be clear in your head about what the hypothesis says and what would falsify it.

As far as I can see from your account of things the hypothesis says ‘Any organ in the body which contains red blood cells will look purple’. It does not say ‘All the purple organs in the body look purple because they contain red blood cells’.

If you find any purple organs in the body which contain red blood cells then this will support the hypothesis. If you find purple organs in the body that do not contain red blood cells then these organs will not support the hypothesis but they won’t falsify it either.

So what will falsify the hypothesis? If you find an organ in the body which clearly looks green or orange but which contains red blood cell then this would falsify the hypothesis.

You need to be clear in your head about two similar looking hypotheses which are logically very different.

1. All organs with red blood cells in them will look purple 2. All organs which look purple will contain red blood cells.

Your hypothesis is 1 not 2 So if you find a purple organ with no red blood cells in it. This will NOT falsify 1 but it will falsify 2.

If you find a green organ with red blood cells in it then this will falsify 1.

Answer by Craig Skinner

Your students might learn some interesting zoology, but they wont learn much about hypothesis testing from this exercise.

The ‘root of the mess’ isn’t that toads have more than one (externally) purple organ.That’s an interesting fact for your students to learn. Nor is it that your hypothesis isn’t falsifiable.

The problem is that you don’t have a clear hypothesis to test. Your ‘hypothesis’ can be construed either as a conditional (if….then…) or as an argument (RBCs contain Hb, which is purple; therefore RBCs are purple; therefore an organ storing RBCs will appear purple, internally and maybe externally). But it isn’t a hypothesis (conjecture).

In the light of what we believe about RBCs and Hb, and what we surmise about the RBC life-cycle including a storage phase, a hypothesis (conjecture) for testing might be:

‘The toad organ-set includes at least one which appears purple.’

(falsified if we don’t find one – maybe because 1. RBCs stored in bone marrow, bones appear white, only insides are purple, or 2. No RBC storage in toads, or 3. Stored RBCs change their colour)

In fact we find several purple organs (say spleen, liver, two kidneys). So hypothesis confirmed (for one or a few toads at any rate, we surmise most others are the same). One, or more, or none of these organs might store RBCs. So, a better, more specific hypothesis might be:

‘The toad organ-set includes at least one which appears purple due to stored RBCs.’

(falsified if either no purple organ or none contains stored RBC when examined microscopically) In fact we find spleen contains large numbers of RBCs but liver/kidneys contain other things and only a few RBCs inside their blood vessels, as with any organ. Hypothesis confirmed (for those toads examined at any rate)

I don’t think induction comes into your conjecture-and-testing exercise with toads, except that the move from ‘all toads examined have a spleen’ to ‘all toads have a spleen’ is inductive. Of course a few toads may well be asplenic, so the better inductive argument would be to ‘nearly all toads have a spleen.’

A personal note. I don’t mind carving up toad, human or any other sort of dead bodies. It’s the killing of them for this very purpose which I am unhappy about, even with toads. But that’s another debate.