|Page tools: Print Page Print All|
When discussing issues in the development of questionnaires, Richard Platek, Director of Statistics Canada, underlined a basic truth in form design:
'The subjects of measurement are human beings and the process of measurement is based on language.'
If, through inappropriate language, the respondent cannot fully understand what we are asking, he or she cannot provide what we want. In short, because we are dealing with people and the meaning of words, the potential for misunderstanding is enormous.
All this translates into a 'we have to make them understand' view of respondents. That view has two components: the 'them' and the 'understand'. This section addresses the language that connects these two.
First, the 'them': who are these people who complete ABS forms? They are Australians of all social strata and ability levels. They are not blessed with some special insight into statistical language just because we sent them a form. While the ability of respondents to understand questions and explanations varies, it is all too easy to overestimate this level of ability.
David Sless, Director of the Communication Research Institute of Australia, suggests that forms designers should work with the following minimal expectations of respondents:
There is a myth in the ABS that respondents understand our definitions and jargon, and that our explanations are well understood. The evidence suggests that is not generally the case.
For example, ask a small business respondent what the difference is between establishments and type of activity units? Chances are the respondent will have some difficulty in understanding these words, which are part of the ABS language, but not part of the respondent's.
Words with precise definitions are an integral part of ABS forms. However, unless evidence from tests and evaluations shows that respondents understand the meaning, we must make the meaning clear in appropriate language. We cannot presume every respondent speaks our language. Obviously not every respondent will understand the meaning of 'bailment', or 'leverage lease agreement' for example. Use appropriate language to explain the meaning, or find a form of words which is clear to your respondents.
Don't be concerned about 'talking down' or insulting the intelligence of respondents through simple language. Nobody is insulted by something they can understand easily.
If, through testing, you know your group of respondents understands 'bailment' then you can include the word in the form with confidence.
In essence, the way to make 'them' understand is to talk to 'them' in the language they understand. ABS officers often find themselves explaining forms to respondents on the telephone in conversational appropriate language. Used in this form, appropriate language short circuits the need for additional explanations. The principles of appropriate language are detailed in the following paragraphs.
Briefly, the title of the form should:
This is particularly the case where forms are sent to people or organisations who receive a large number of forms, and yours has to be readily identifiable amongst these others. It is also a matter of courtesy to let the respondent know straight away what is being asked.
A form title ideally consists of only two parts: a subject (2004-05 Economic Activity) and a function (survey), e.g. Economic Activity Survey 2004-05.
The title is not a sentence, and should be brief and meaningful.
Evidence suggests that a section heading often prompts a respondent to skip over a section if it appears to not apply to their situation. e.g. the 'Drivers under 25' section on NRMA car insurance proposals applies to all applicants but this section was often ignored by those over 25 years of age. The amount of follow up required was dramatically reduced when the section heading was removed.
Section headings can be useful where a respondent is directed to skip over sections of a form. In this situation section headings can help in the quick identification of the relevant sections of a form.
There are two main types of instructions used on forms. The first type deals with what to do with the form, where to get help and so on.
The second type deals with how to answer questions.
Instructions are frequently ignored on forms. Instructions placed at the start of forms, if read at all, are often forgotten by the time the person has answered one or two questions.
Short-term memory is used extensively when completing forms. This means that instructions need to be placed as closely as possible to the place where they are to be applied. This often means repeating the same instruction each time it is to be acted upon. This is particularly the case with instructions on how to answer questions.
As mentioned earlier, the task of completing a form is not regarded as a reading activity. On receiving a form the first thing a respondent is likely to do is grab a pen and write something in the first available space. Instructions might only be referred to by them if they get stuck.
Instructions as to what to do with the form are best grouped together, either at the very start or the very end of the form. The ABS has adopted a standard of putting these all on the first two pages. The front page also contains information about why the information is being sought, the authority for the collection of the data, and names and phone numbers of where to get help.
Instructions at the end of a form often include a checklist of items from the form and details of what to do next.
Short sentences are easier to understand than long ones. Although it is unrealistic to specify a maximum length, each sentence should convey a single item of information.
A poor sentence combining many items of information is shown in the following example (see Diagram 7.1).
Breaking the instruction into sentences containing one idea each and improving the layout gives the following presentation (see Diagram 7.2).
3 Report all stocks owned by the business named on page 1
Order of clauses
Sentences should have clauses in chronological order to aid comprehension. In the example below (see Diagram 7.3), the first action is to read the instructions, but this doesn't become clear until the end of the sentence.
A better method is shown below (see Diagram 7.4).
Respondent / activity relationship
Keep clauses naming the respondent close to the clause which says what the respondent has to do. The example below (see Diagram 7.5) inserts an explanatory phrase before the action is complete.
The following example (see Diagram 7.6) is a clearer way of giving the same instruction.
Positive versus negative questions
Research has shown that when people are asked negative questions they are much slower in responding. The slower response is caused by people having to think about the answers rather than giving them automatically. Although we might consider that causing people to think about questions is good, with many questions people think wrongly and give the wrong answers. Negative questions cause additional effort for our respondents, for no good reason.
The following is an example of negative instruction (see Diagram 7.7).
The positive alternative is shown below (see Diagram 7.8) where the instruction 'tick one only' is positive and unambiguous.
Active versus passive voice
Most people find sentences in the active voice easier to understand than the passive voice.
Passive voice is when the subject is acted upon, for example: "This form is to be completed by the operator whether operating for the full year or only part of the year."
Active voice is when the subject performs the action: "The operator is to complete the form, whether operating for the full year or only part of the year."
Making nouns out of verbs
This is a common practice with many writers. In the following example the verb 'certify' is made into a noun: "Certification for all collection forms must be given by an Assistant Statistician or Deputy Australian Statistician."
The preferred alternative is much clearer, uses fewer words and is more direct: "An Assistant Statistician or Deputy Australian Statistician must certify all collection forms."
Form designers frequently try to save space on a form by omitting phrases that they consider to be unnecessary but which would have helped clarify the meaning.
The following example shows a sentence which has been abbreviated by leaving out some of the words which would be used in spoken English: "Comment here on unusual circumstances (drought, floods, fires, hailstorms etc.)."
The conversational style of the next example is more easily understood by people with an average reading ability:
Words of more than two syllables should be avoided as much as possible, although length alone should not be the criterion for evaluating their use. Some short words such as redress are not familiar to most people, while other multi-syllable words may be very well known. This does not mean that long words should never be used. However, they are frequently misunderstood by poor readers and this should be taken into consideration in the design. Only an evaluation of the readability of the form e.g. using observational studies or interviews with respondents, will provide the answer.
The use of words in a particular context may change the meaning. A word when used alone may be quite familiar to people, but when used in a specialised context it may not be familiar. In many cases the reader will have a clearer concept of what is to be filled in if the reasons for wanting the information are known. ABS forms may carry a brief description on the use of the statistics collected. Apart from any legal obligations, respondents may require an indication of why they are required to complete the form, and what ABS will do with the information collected.
Technical and statistical terms are to be avoided unless you have evidence that respondents understand what they mean. If you do not have this evidence, explain the term in appropriate language. Again, the question here is, do the respondents understand what is meant by technical terms and industry-specific definitions? For example 'leased assets', according to an ABS definition, refers to 'fixed tangible assets leased/hired under a finance lease or leverage lease agreement which effectively transfers from the lessor to the lessee substantially all the risks and benefits incident to ownership'. In short, if technical terms are used, explain them and test them to be sure.
Punctuation aids comprehension if used correctly. Only commas, full stops and question marks are readily understood by most people. The burning need for a semicolon or a long succession of commas probably means a new sentence is needed. Further information is provided in the Australian Government Style Manual.
Just as beauty is in the eye of the beholder, words are what the reader makes them. The meaning ABS attaches to the words may not be the meaning the respondent attaches to them. The only way to find out is to test the words on the respondents.The principles of appropriate language are a solid foundation on which to base our approach to the language used in forms. The principles are not a substitute for testing particular words with particular groups of respondents.