Ten guidelines for creating and using hypotheses Here is how to make a useful research hypothesis: STEP 1: THE QUESTION
Our example starts with a small article in a journal for chemists. Scientists at the Technical University of Copenhagen had a test done with plastic pipes for drinking water. They had filled tubes with water in their laboratory and left it there for a couple of days. Then they had run the water through a so-called ‘gas chromatograph and mass spectrometer’, to see if maybe chemicals from the plastic tubes had migrated into the water. To their surprise they found in the water not one or two, but more than twenty chemicals that only could have come from the plastic. Two substances they could immediately identify: phenols, and plasticizers. The other substances were unfamiliar. Whether they constitute a danger to health, the researchers could not say. But they knew one thing for sure: neither of the chemicals belongs in drinking water. ‘We thought it a useful exercise,’ said the initiator of the study, Erik Arvin, when we visited him in his laboratory. ‘Whenever a new material is applied, problems occur. So we wondered whether the chemicals in the plastic could leak into drinking water.’ Arvin and his colleagues gave us a great subject on a platter: Something quite normal - drinking water – but potentially very dangerous. (By ‘we’ I mean: Kaare Gotfredsen, Marleen Teugels, Chris Vermeire and I, who in 2005 worked together on the investigation of the water pipes. We published our story in the Belgian magazine Knack, the Dutch newspaper Algemeen Dagblad and the Danish newspaper Fyen Stift Tidende.) When we returned from our first visit to Copenhagen, one question echoed in our minds: Are plastic water pipes, which are so often used in new kitchens and bathrooms, a danger to public health? STEP 2: THE ANSWER We answered the question ourselves - before we started up Google or put on our coats for an interview. We did not want to be influenced, at this stage, by the opinions of others. We tried to answer the question we had asked (Are plastic water pipes a hazard to health?). We wanted the answer to be based on our own knowledge and experience. Such an answer is called a "hypothesis". A hypothesis is a proposition, an assumption, an idea. The hypothesis must be proved in practice. It is even likely that one will have to adjust an hypothesis several times (see below). But scientists, investigators and other researchers know from experience that it is easier to examine a hypothetical answer, than a question. Therefore: Rule 1: Formulate your hypothesis as an answer to the question: “What is going on?”
Copyright 2010 Luuk Sengers Journalistiek | www.luuksengers.nl 1
The hypothesis that first came to us was: People get ill after drinking water from plastic pipes. Not coincidentally this sounds like the header of an article – as journalists, we tend to formulate strong headlines. But as researchers we should put higher demands on the formulation of our hypotheses. The assumption that people get ill by drinking water from plastic pipes sounds catchy, but then immediately demands: How does it actually happen? Who exactlyis affected? How ill do they get? The purpose of a hypothesis is not to ask questions, but to answer them (see previous step). In proposing prospective answers to all the questions implicit in our hypothesis, we automatically make the hypothesis more concrete. What we should do, first and for all, is introduce a recognizable main chararacter. One that can be hold responsable for the act that we are investigating. By choosing an active protagonist, we force ourselves to designate a “perpetrator” (negative) or “hero” (positive) in our story. There are not many better ways to make a problem more concrete! In our example, persons getting ill are the main characters. But they don’t do anything, do they? Getting ill just happens to them. Looking for an active protagonist, we came to this hypothesis: Pipe manufacturers add chemicals to their plastic water pipes which later mix with water, making people ill. So that makes
Rule 2: Choose an active protagonist for your hypothesis.
Let’s do some more filling in. After reviewing the scientific article of the Danish chemists, with which our investigation began, we could more precisely formulate our hypothesis as follows: Pipe manufacturers add plasticizers and phenols to their polyethylene pipes which later mix with water, making people ill.
Indeed, for the headline above a news article this is way too long and too complicated. But remember: a hypothesis is a working tool and not a finished product. The purpose of a hypothesis is to describe an event as specifically and in as much detail as possible, in order to make it easier to investigate. The scientific article stated that above all, plasticizers and phenols were discovered in the pipes. And only in pipes made of polyethylene (one of many types of plastic). What the chemists could not say, was if there was any danger to public health. Hence our vague notions “people” and “sick” in the hypothesis. We had to hang on to them until we could get more information from other sources. Rule 3: Extend your hypothesis with the answers to the questions: who, what, where, when, how and why.
Copyright 2010 Luuk Sengers Journalistiek | www.luuksengers.nl 2
Our hypothesis: Pipe manufacturers add plasticizers and phenols to their polyethylene pipes which later mix with water, making people ill is probably concrete, but not very easy to remember or work with. It is long and complicated. A dragon like this could seriously limit your passion for the job ahead. To restore the overview, you can cut the hypothesis in smaller steps, called “partial hypotheses”. Each part is an assumption in itself of an isolatetod event - which of course again has to be formulated as concretely as possible! When the hypothesis is divided into smaller, individual steps, it will also reflect reality. In reality, events usually are built of a sequence of causes and consequences. Moreover, splitting up the hypothesis forces you also to identify multiple “perpetrators” and / or “heroes”, since every partial hypothesis has hits own protagonist. Again, this comes closer to reality, in which rarely one person alone is responsible for a great event. This is how we split up our hypothesis, and supplemented it with new details that we discovered in our investigation: 1. Pipe manufacturers add plasticizers and phenols to their polyethylene pipes, because they are cheaper than harmless alternatives.
2. Installers and do-it-yourselfers use polyethylene pipes in homes and offices on a large scale, because they are easy to use.
3. The phenols and plasticizers in the pipes intermingle with water when it stands still for a couple of days at room temperature.
4. The phenols and plasticizers in drinking water cause cancer and infertility in humans, because they mimic the female hormone estrogen, thereby disrupting the endocrine system.
This is easier to handle, isn’t it? Hence: Rule 4: Cut your hypothesis into smaller, sequential events. The problem with the last two steps is that we cannot see them. And what you cannot see, is often hard to prove. One step takes place in a closed pipe, the other in a human body. Moreover, these are microscopic substances and not visible to the naked eye. Therefore: try to formulate your hypothesis as a stage or film direction – as an event that can be acted out before a camera or an audience. This is not only a useful tip for TV and radio journalists who need images and sounds. Even print journalists make it easier for themselves – if they want to present the news in a story – if they formulate their partial hypotheses as a kind of scenario (see below). But the main reason for choosing visible events is that they are easier to prove! We replaced partial hypothesis 3 by:
Copyright 2010 Luuk Sengers Journalistiek | www.luuksengers.nl 3
Scientists pour water into polyethylene pipes. They put the pipes in a laboratory room and turn the thermostat to 23 degrees. After seven days they measure the chemicals in the water with a special device. They see in the results mainly phenols and plasticizers. Just by swapping matter (phenols and plasticizers) for people (scientists), we have solved the problem. Changing the fourth partial hypothesis (effects on health) into something visible was less easy. From a journalistic point of view (but not from a human point of view!) the following partial hypothesis had been wonderful: People drop dead after drinking water from pipes of polyethylene. But there is (fortunately!) no evidence for that. Getting ill from chemicals is a much slower process, apparently. And, as often in cases of cancer and infertility, the cause might not be reduced to just one culprit. Were the effects of plasticizers and phenols on human health ever tested? we wondered. In a controlled environment that excludes other influences? If ever such tests were done, then they were obviously done not people, but on animals. That brought us to the following hypothesis: Scientists administer phenols or plasticizers to mice. Then they examine the mice. They see in the results signs of cancer and / or infertility. Purely by the exact wording we were, after some tracking on the Internet, able to establish the link that we were looking for: that the phenols in drinking water pipes are very harmful to human health! By typing the words: “scientist, research, cancer, phenol” into Google, we finally found a treatise on scientists who, totally by coincidence, had found in their laboratory a link between plastic and cancer. And in a very bizarre way. Cell biologist Ana Soto told us in an interview: “I was with two colleagues engaged in a study on the impact of estrogen – female hormones – on the development of breast cancer. We knew that the cancer cells began to proliferate in test tubes when they were brought into contact with estrogen. But to our surprise exactly the same happened in the control tubes, to which were added no estrogen. Incomprehensible!” “Like Sherlock Holmes we went looking for the perpetrator,” Soto recalls. “Meticulously, we studied all testing phases and checked all equipment. Only after months of investigating, we discovered the cause: The estrogens originated from the plastic of the test tubes. The plastic contained nonyl-phenols that appeared to mimic female hormones.” Soto would in the following years develop into the world’s leading researcher in the field of cancer and phenols. Because we created an internet “alert” for the keywords “scientist”, “research”, “cancer” and “phenol”, we were automatically notified of new articles on the web. So a few months later we were “alarmed” about another U.S. study. Mothers with a high concentration of plasticizers in their urine during pregnancy, gave remarkably often birth to boys with stunted penises and other genital abnormalities, making it difficult for them in later life to have children. For the first time, with this study, a direct link was established between hormone disrupting substances and reproductivity problems in humans.
Copyright 2010 Luuk Sengers Journalistiek | www.luuksengers.nl 4
Rule 5: Formulate your hypothesis as a stage or film direction – as an event that can be acted out before a camera or an audience. Is drinking water not monitored, you may ask? We too wanted to know. According to the government, public drinking water is one of the most scrutinized food products. How, then, can such hazardous substances appear in water intended for consumption? This is what we thought might be the reason: The government does not monitor drinking water for phenols or plasticizers.
With this hypothesis, however, a problem emerges: how do you prove something that does not happen? What is not there, can not be shown. There is a number of ways to solve this problem: 1. Turn the argument around and show what the protagonist does do. For example: The government monitors drinking water for other chemicals than phenols and plasticizers.
Or: The government monitors drinking water on the basis of a list of substances that does not contain phenols and plasticizers.
2. Let the protagonist prove that he does not do something. Like in one of these two ways: The government says it does not check drinking water for phenols and plasticizers, because the process is expensive.
The government notes in its guidelines that drinking water does not have to be checked for phenols and plasticizers. 3. Let others prove that something does not happen. For example: Experts say that the government monitors drinking water with equipment that is not suitable for detecting phenols and plasticizers.
Rule 6: Do not tell what is not happening. Don’t chase after ghosts. By the way: now that we are in the process of adding new partial hypotheses to the investigation, let’s not forget this one: so far we identified the causes and consequences of the problem (phenols and plasticizers in drinking water), but not the solutions. We think we do our readers a favor if we do not just warn them, but also offer them a way out. How can they avoid the hazardous chemicals in drinking water pipes? While we were doing our research, the European Parliament prepared a new directive (= law)
Copyright 2010 Luuk Sengers Journalistiek | www.luuksengers.nl 5
for dangerous substances. Leading scientists from around the world, who worried about the health risks of phenols in plastic, lobbied in Brussels to put phenols on a black list. A prohibition on the use of phenols in consumer products was in their opinion the only solution. That brought us to the next, final step in the process: Scientists put pressure on European parliamentarians to ban the use of phenols and plasticizers in consumer products.
STEP 3: SCENES
The best moments are when you are able to “translate” your partial hypotheses in scenes. A scene is more than a “normal” event: it is much more exciting! First, a scene features a main character (and not an anonymous group). And, more importantly: in a scene the protagonist encounters an obstacle. He faces resistance that he must overcome to be successful. There are potentially strong scenes in the partial hypotheses above. Look what happens when we formulate them slightly differently (and note the word “but.”): Erik Arvin pours water into polyethylene pipes. He then puts the pipes in a laboratory room and turns the thermostat to 23 degrees. After seven days he measures the chemicals in the water with a special device. He sees in the results mainly phenols and plasticizers. But there are other substances he does not recognize and he cannot rule out with certainty that these are harmful for humans after consumption.
Ana Soto expects the cancer cells in the control tubes without estrogen not to grow, but they do.
Sixty prominent scientists, including Ana Soto, published a manifesto in which they call the European Parliament to ban the use of phenols and plasticizers in consumer products. But ultimately, the European Parliament allows the use of phenols anyway.
Rule 7: Turn your hypothesis into a scene by entering an obstacle, a “but…”.
STEP 4: THE RESEARCH
Now go and put your hypothesis to the test. Look for convincing evidence that your assumption is correct. The chance that your first hypothesis is complete and spot-on is not big. At the beginning of your research you simply lack enough information to draw an accurate picture of what happened. With the facts you encounter in interviews and documents, you can improve your hypothesis. Sometimes these are additions or details, but other times you must also change the direction of your hypothesis, because things turn out to be just a little different than you thought. By recrafting your hypothesis, you increase the probability it is correct. Gradually the assumption turns into a statement. We discovered in our research, for example, that it is not entirely the fault of the pipe manufacturers that phenols seep into drinking water. Some phenols appear to arise only by a
Copyright 2010 Luuk Sengers Journalistiek | www.luuksengers.nl 6
reaction of the chemicals in the plastic with the water. This meant a considerable weakening of our first hypothesis regarding the role of the pipe manufacturers (Pipe manufacturers add plasticizers and phenols to their polyethylene pipes, because they are cheaper than harmless alternatives). At other points we could strengthen our hypothesis: other research on plastic water pipes, for example, showed that the danger not only comes from polyethylene pipes, but also from much more widely used PVC pipes. Rule 8: Adapt your hypothesis to the facts - and not the other way!
An hypothesis gives direction to your research. It’s like a signpost. By following it, you avoid detours and getting lost. And that saves time. To save yourself even more time, you can set a lower limit to your hypothesis: how far are you willing to weaken your hypothesis before calling off your research? In other words: with what hypothesis do you still have a story? Such a limit is called a “minimum” hypothesis. If you cannot prove this minimum hypothesis, then you quit your inquiry. If you start creating an hypothesis, you first reach for the stars: you want the biggest news possible. In our example of drinking water that is: People drop dead after drinking water from plastic pipes.
It soon became clear, however, that we would never be able to prove that. So we had to consider the question: how far will we weaken or hypothesis? Do we still have a story when people drinking water from plastic pipes ‘only’ get sterile? And what if we only find that people can get a serious cold from drinking out of plastic pipes? Or if it appears that only very old people are at risk? Or if we find no conclusive evidence that the chemicals can be bad for health? Or is just that fact - that it can not be proven that the substances are harmless - reason enough for a news story? We chose the latter. It was therefore that we formulated our minimal hypothesis as: The government says it cannot guarantee the safety of plastic drinking water pipes.
Rule 9: Formulate a minimal hypothesis - the minimum you need for a news story.
Finally a tip: start your investigation with the partial hypothesis that is easiest to prove. Go in through a door that is already halfway open. It is simpler, and even if this hypothesis proves to be false, you at least know that you don’t have to investigate any further. In case the partial hypothesis turns out to be true, then that is a huge encouragement to tackle the more difficult partial hypotheses next. In our example this appeared to be the easiest partial hypothesis: 2. Installers and do-it-yourselfers use polyethylene pipes in homes and offices on a large scale, because these are easily to install.
Copyright 2010 Luuk Sengers Journalistiek | www.luuksengers.nl 7
And, indeed, one phone call was enough to receive an email with the sales results of polyethylene pipes, compared with pipes from other materials. Rule 10: Start with the partial hypothesis that is easiest to prove.
For more on the art of hypothesis making, read:
Mark Lee Hunter and others: Story-Based Inquiry: A Manual for Investigative Journalists, UNESCO 2009. Free to download at:
(With many thanks to Mark Lee Hunter, who reviewed this paper and added his valluable advise.)
Copyright 2010 Luuk Sengers Journalistiek | www.luuksengers.nl 8
St. Johann Baptist – Normannenstr. 73, 42277 Wuppertal, Telefon 660433 St. Konrad – Hatzfelder Str. 263, 42281 Wuppertal, Telefon 524262 St. Maria Himmelfahrt – Wittener Str. 75, 42279 Wuppertal, Telefon 660055 St. Marien – Amselstraße, 42281 Wuppertal, Telefon 505236 St. Pius X. – Liebigstr. 7, 42283
Kurztherapie mit sozial und kulturell benachteiligten Menschen Zusammenfassung Diejenigen von uns, die mit wirtschaftlich und kulturell benachteiligtenMenschen therapeutisch arbeiten, sollten bedenken, daß die Therapeutenrolledirekt oder indirekt großen Einfluß auf die Klienten und Klientinnen erlaubt. Dieser Einfluß sollte zu deren Wohl genutzt werden. Psychotherapie ist einkulturel