differ by age, sex, or disease severity, this could account for apparent differences in improvement between the groups.
This is a weak experimental study. Introduce matching.
1 Match patients in pairs for age, sex, disease severity; assign the first patient in each pair to receive treatment, the second patient to receive a placebo.
The person assigning patients may have a subconscious preference for putting one patient first in each pair. Does the patient know which treatment they are getting?
This is a weak placebo‐controlled intervention trial. Introduce randomization and blinding.
1 Allocate patients to treatment or placebo randomly within pairs. Make sure that the researcher does not know which patient is to receive which treatment (the researcher is then said to be ‘blind’ to the allocation of treatment). Make sure that the patient does not know which treatment they are receiving (keep the patient ‘blind’ as well). This makes the study ‘double blind’.
2 Conduct a placebo‐controlled randomized double‐blind intervention trial.
‘Randomization properly carried out is the key to success.’
(Sir Ronald Fisher)
Intervention studies of this type are often regarded as the most robust for testing a hypothesis. But sometimes randomized controlled trials are not ethical, especially if the exposure may be harmful (e.g. smoking, or increasing someone’s saturated fatty acid intake), or it is not possible to blind either the subjects or the researchers because the treatment being given is so obviously different from the placebo. In these cases, it is possible to mimic intervention studies in samples that are measured at baseline and then followed up over months or years. These types of studies raise issues about how to deal with factors that cannot be controlled for but which might affect the outcome (e.g. in a study looking at the impact of diet on risk of cardiovascular disease, the influence of smoking and diabetes) and changes in the general environment (e.g. indoor smoking bans) that have the potential to affect all the subjects in the study. Complex statistical analysis can be used to cope with some of these design issues. The ability to test the hypothesis in a robust way remains.
1.4.5 Statistics
The Dodecahedron: ‘Just because you have a choice, it doesn’t mean that any of them has to be right’.
Statistical tests enable you to analyze data in order to decide whether or not it is sensible to accept your hypothesis. There are literally thousands of values that can be calculated from hundreds of tests, but unless you know which test to choose, the values that you calculate may not be appropriate or meaningful. One of the main aims of this book is to help you learn to choose the test which is right for the given research problem. Once you have decided which test is appropriate for your data, the calculation is a straightforward manipulation of numbers. It is vitally important, however, to learn which data to use, how the manipulation is carried out, and how it relates to the theoretical basis which will enable you to make the decision about the truth of your hypothesis.
Most of the time it is better to use a computer to do the computation for you. Even if you enter the values correctly and generate a meaningful outcome, the computer will not tell you if your hypothesis is true. For that, you need to know how to interpret the results of the tests.
1.4.6 Interpretation
The Dodecahedron: ‘If you want sense, you’ll have to make it yourself’.
Every statistical test will produce a number (the test statistic) which you then need to interpret. This is the last stage and often the most difficult part of statistical analysis. The final emphasis in every chapter that deals with statistical tests will be on how to interpret the test statistic. We will also look at the SPSS output to verify that the right set of values has been entered for statistical analysis.
Two concepts deserve mention here: ‘Inference’ and ‘Acceptance’. ‘Inference’ implies greater or lesser strength of fact. It is usually expressed as a probability of a given result being observed. If there is a high probability that the result which you have observed is associated with the hypothesis being true, we talk about ‘strong’ evidence. If the observed outcome is little different from what we would expect to see if the null hypothesis were true, we talk about ‘weak’ or ‘no’ evidence.
At some point, we need to make a decision about whether to accept or reject the null hypothesis, that is, to make a statement about whether or not we believe that the hypothesis is true. ‘Acceptance’ implies a cut‐off point upon which action will be taken. We will discuss cut‐off points in Chapter 5. It is important not to confuse political expediency (acceptance) with scientific validity (inference).
1.5 NEXT STEPS
Every year, at least one student shows up at my door, holds out an open notebook with a page full of numbers, and says, ‘I’ve collected all this data9 and now I don’t know what to do with it’. I strongly resist the temptation to tell them to go away, or to ask why they didn’t come to see me months ago. I usher them in and see what we can salvage. Usually, it is a debacle. The data collected are not suitable for testing the hypothesis; their sample is poorly defined; they don’t have enough of the right types of observations; they have used different methods for collecting data at baseline and follow‐up; the list goes on and on.
Box 1.3 summarizes the steps that should be undertaken when conducting research. Although Steps 1 and 2 are essential (‘Getting the question right’), probably the most important step is Step 3, the point at which you design the research project. It is vital at this stage that you consult a statistician (as well as others who have done similar research). Be prepared to accept that your hypothesis may need modifying, and that the design that you first thought of is not perfect and would benefit from improvements. It is very unlikely that you will have got it right at your first attempt. Be prepared to listen and to learn from your mistakes. As I said in the Introduction to this book, statisticians may be perceived as monstrous, inhuman creatures intent only on humiliating those who come to consult them. In reality, the statistician is there to advise you concerning the likelihood of being able to prove your hypothesis, guide you in the design of the study, the choice of measurements which you intend to make, and the type of analyses you plan to undertake. Months or years of effort can be wasted if you embark on a study which is flawed in its design. Do not take the chance! Be brave! Be thick‐skinned! Talk with statisticians and accept their advice. Even get a second opinion if you feel very uncertain about the advice you are given.
1.6 RESEARCH DESIGN
There is a wide variety of research designs which can be used to address the many research questions that you are likely to ask. There is no strictly right or wrong answer concerning which design to use. You should recognize, however, that some designs are stronger when it comes to arguing the truth of your hypothesis. The aim in carrying out any research will always be to obtain the maximum information from a given design in relation to a particular research question, given the time and financial resources that are available.
1.6.1 Project Aims
Coming up with an interesting and useful research question
