
Mastering exam technique
Peter Cantillon gives some tips on dealing with MCQs and SAQs
The multiple choice questionnaire (MCQ) has become one of the standard assessment methods in most medical schools over the past 30 years. MCQs were developed to address problems associated with traditional essay style testing. They are seen as a more objective test of students' knowledge than essays. This is because essays are marked by individuals who interpret a student's response against a standard answer format. So there is inevitably an element of personal judgment or subjectivity about such a marking scheme. MCQs, on the other hand, should have clear right and wrong answers and so should be free of any interpretation bias. They are generally regarded as being reliable and thus consistent and valid tests of students' knowledge and information recall.
MCQs have many advantages for teachers. Once designed they are easy to administer and mark. Most MCQs are answered
on a card, which can be read by a computer coupled with an optical card reader.
So the whole process can be automated
leaving teachers free to do other things.
The scoring is done by the computer,
which also reports on the reliability and
discriminatory power of the MCQ questions. The crude unadjusted results are
available quickly and can be given to examiners who are in the process of judging
pass or fail and honours performance. So
it seems that the MCQ is an assessment
design that you are likely to encounter
many times during your undergraduate
and postgraduate career.
It is important to note, however, what
MCQs cannot do. They cannot measure
clinical performance or competence. They
can occasionally be used to test components of diagnostic reasoning and problem
solving.
They are usually combined with other
test formats that test different aspects of
candidates' performance - for example,
objective structured clinical examinations,
short answer questions, etc. MCQs have
also been criticised for providing candidates with visual "cues" that suggest a correct answer, one which the candidate might not otherwise have recalled.
Marking schemes
There are two types of marking commonly encountered in MCQs. These are "negative marking" where marks are deducted for incorrect answers and "neutral marking" where all questions should be attempted - marks are given for correct answers only and guessing is encouraged.
Negative marking was introduced to
reduce the amount of guessing by candidates by "punishing" incorrect responses
with a negative mark. There are several different MCQ formats that may be encountered in medical school examinations, the commonest of which is the true/false design (see box 1).
| Box 1: True/false design
In a 67 year old man with mild hypertension, first line antihypertensive treatment would usually be started with:
| Clonidine |
T |
F |
DK |
| Atenolol |
T |
F |
DK |
| Bendrofluazide |
T |
F |
DK |
| Methyldopa |
T |
F |
DK |
| Amiloride |
T |
F |
DK |
|
The question begins with a stem or
question and then poses five solutions.
Each solution may be true or false. Some
examinations also offer the candidate a DK
(don't know) option.
An alternative design (simple multiple
choice) is where only one of the solutions
is the correct answer (see box 2); the other
four are designed to distract you and are
in fact called "distractors." This design
reduces the likelihood of success in guessing. In a standard true/false design you
have a 50% chance of being right if you
guess, whereas this method means you
have only a 20% chance of guessing the
right answer.
| Box 2: Simple mutliple choice
The drug of first choice for the management of hypertension in a patient with diabetes who is insulin dependent with no pre-existing renal or cardiovascular disease is:
| Bendrofluazide |
T |
F |
| Atenolol |
T |
F |
| Clonidine |
T |
F |
| Frusemide |
T |
F |
| Enalapril |
T |
F |
An angiotensin converting enzyme inhibitor would be regarded as the drug of first choice from this list; the other drug options act as distractors.
|
The stem at the top of an MCQ question can take many forms. It may be a
statement or question. It could also be a
brief case description, a set of results, an
electrocardiogram strip, or even a photograph.
A common MCQ format used in the
United States, which is beginning to be
used elsewhere, is the "extended matching"
design. This purports to allow testing of
higher level functions such as diagnostic
skills. The candidate is offered a long list
of answers and has to match them to a
short list of clinical problems or scenarios.
MCQ tips
Preparation
Try to gather as many examples as you can
of old papers and previous examples of
MCQs used by the department or school
in question in the past.
Do not, however, try to memorise hundreds of responses to questions. The factual knowledge you will gain will be superficial and dissociated. It is better to look for the topic areas that recur frequently and ensure that you have a deeper knowledge of these topics.
Revise with friends and colleagues. You
can share knowledge and techniques.
Familiarise yourself with the optical
reader cards that you will be using to
record your answers in the exam.
Examples should be available from the
"examinations office."
You should know what type of MCQ is
being set for you. Will there be negative
marking? How much time will you have
and how many questions will there be?
On the day
Check that your understanding of the
MCQ format is correct. "It is negative
marking, there are 300 questions, and I
have two hours to complete this."
Always read the stem for each question
carefully. Have you understood the question? Are there any ambiguities? If so ask
an invigilator who will alert an examiner.
There are usually one or two in the room.
Allocate three quarters of the time to answering the questions and a period at
the end to checking answers and accuracy.
If the MCQ is a true/false negative
marking design start by answering questions for which you are certain of the correct answer(s). Mark the questions where you may know the answer but need to
think a little. When you have completed
your "certains" then look at the questions about which you are less certain. Answer as
many of these as you can. It is generally advised that
you do not guess in situations where you "haven't a clue." The half and half rule applies both ways. You may improve your position; you could also worsen it considerably.
If the MCQ has a "neutral"
marking design you should answer all
questions. You will not be penalised for
guessing, so have a go. You cannot score if
you leave answers blank.
Essay and short answer question assessments
Essay tests were for many years the standard written assessment in medical schools
throughout the world. They have lost
favour in recent times and have been
replaced by newer formats such as short
answer questions (SAQs), multiple choice
questions (MCQs), and modified essay
questions (MEQs). From a teacher's point
of view essays are easy to set but are notoriously time consuming to mark. It is also
difficult to achieve reliable marking
between different sets of examiners marking the same papers.
The standards of assessment expected
of medical schools are now high, so unreliable and time consuming assessments are
being phased out. Having said that, the
essay has an important place in the constellation of assessment methods. It is a
useful method for assessing students'
capacity for analytic thinking and their ability to present their ideas coherently (and
legibly). Essays can also be used to assess
students' attitudes and their aptitude for
looking at a problem from different angles.
Essays may take several different forms. The commonest styles are
extended response and restricted
response. The extended response question
tests the depth of the students' knowledge
and their ability to organise and present
their thoughts. The restricted type is used
to test the ability to recall knowledge related to a particular topic.
These are two examples.
Describe the community based services
that you believe ought to be provided to
patients after a stroke. Support your
answers by providing evidence of effectiveness for each intervention. (Extended
response essay question)
A 76 year old woman presents with a mass in her right breast. A biopsy confirms cancer. Present your plans for investigation and management. (Restricted response essay question)
Essay tips
Preparation
Try to acquire past papers from the examining department. You may also be able to
get papers from students in the years
ahead of you. There are books of essay
questions available.
Work with friends and colleagues to plan answers to questions. Share knowledge and techniques
Predicting likely essays is a risky business. It is reasonable, however, to try to
identify what you (and the examiners) think are the areas of core knowledge for the essay examination. Check your estimation of key areas with friends and colleagues. You can then devote more time to acquiring a fuller understanding of those core areas while maintaining a strategic knowledge of other apparently peripheral topics. You cannot know it all so you do have to choose where to target most of your efforts.
Doing the essay
Check how much time you have to do the question(s)
that have been set for you. Plan an equal amount
of time for each question. Be disciplined.
Do not be tempted to spend more time on
one question than another. You will get
no marks for an unanswered question no
matter how brilliant the answer to the preceding
question.
It is usually advisable to allocate time for
preparing a framework of how you intend
answering the question before you start to
formally write your answer.
Leave time at the end (15 minutes?) to reread and correct your answers.
If you are feeling uninspired try using surgical and medical sieves to come up with differential diagnoses and other relevant themes. Use categories such as acute and chronic to suggest new ideas. Another useful technique is to organise your essay along the lines of a standard clinical examination - for example, history, examination, investigations, treatment, etc.
Short answer questions
Short answer questions (SAQs) were
designed to address some of the major
criticisms of essay style questions. An
SAQ is designed to sample a student's
knowledge covering a greater proportion
of the content of a course or curriculum
than is possible with in essay. It is also
easier to achieve reliable (consistent)
marking between different examiners
(using standard answers and marking
schedules) than is possible with essay style
questions. The duration of the examination is often similar to that of an essay
style question, but in place of one or two
essay questions there are several questions to which short focused answers are
expected. It is possible to structure the
answers in sequential points rather than a
more discursive essay style (see example
below). The candidate is usually presented with a brief clinical scenario followed by a series of related questions.
The marks for answers to each question
are usually clearly indicated.
An alternative to the standard SAQ that
you may encounter is the modified essay
question (MEQ). The MEQ was designed
to test problem solving and decision
making ability. The MEQ format usually
involves a case history that is sequentially
revealed to the candidate. An MEQ is
usually introduced with a brief case
scenario followed by a short answer style
question. Having completed the first
question the candidate turns to section two
of the problem where a supplementary
question is posed with or without further
clinical information and so on.
Example of an SAQ
A 67 year old woman presents to her GP
after several recent episodes of haemoptysis. She has had a cough for three months
and has lost 6 kg in weight. She has
smoked heavily from the age of 16 years.
Question 1: What is her most likely diagnosis?
Question 2: List three investigations to confirm the diagnosis.
Question 3: List the therapeutic options for the different tumour types.
SAQ tips
As with all written examinations it is important to try to collect past papers. The best preparation is to practise the SAQ format with friends and assess each other's performance.
Get to know how different marks are
allocated for each section within a question.
Try timing yourself doing a past test. See
whether you are able to distribute your time
effectively between the questions. Once
again you must plan to be disciplined about
devoting equal time to questions. You can
spend more time on subsets of questions
that yield more marks, but the overall time
allocated per question should reflect the
marks the question is worth. If all questions are scored the same then the time spent on each question
should reflect that.
During the examination
Read through the paper briefly to acquaint yourself with the task ahead.
Note how marks are allocated. If all
questions have equal weighting in terms of
marks, allocate time evenly. If, on the other
hand, marks are allocated differentially
between questions then devote more time
to questions that yield more marks.
Try using a short point format rather than a long hand style. It saves time.
Recommended reading
Bullimore D. Study skills and tomorrow's doctors. London: WB Saunders, 1998.
The PASTEST series includes useful books giving examples of OSCEs, MCQs, and SAQs. There is little attention, however, to understanding how the assessment works or to exam technique.
Peter Cantillon assistant postgraduate dean
Guy's, King's, and St Thomas's School of Medicine

|