5
Applied Ergonomics 1979, 10.3, 165-169 The role of spatial and typographic cues in the layout ot: journal references James Hartley*, Mark Trueman* and Peter Burnhill t *Department of Psychology, University of Keele. tDesign Department, Stafford College of Further Education. Fifteen American and 15 British readers used the method of paired comparisons to assess 12 different ways of presenting reference lists in social science journals. The results indicated -- despite British preferences for British styles - that the spatial arrangements of elements in the list was the major determinant of preference, and that the presence of a typographic cue only enhanced a preference for a particular layout. Many "kinds of text contain lists of items, each one of which mighl contain a number of sub-elements. Some lists are simple, such as contents pages in books, and some are more complex, such as lists of references or bibliographies. !n deciding how best to present such materials typographic designers have two related problems: first, how to make lists easy ,'o use, and second how, at the same time, to present material economically. Such problems are usually overcome by the judicious use off typographic cueing (ie, the use of capitals, bold-faces, italics, etc). There has been little research, however, into the effectiveness of different ways of presenting the same list and virtually no research at all into the cost-effectiveness of, or the preferences for, different resolutions of the same problem. These latter issues are examined in this paper with reference to designing reference lists for articles in social science journals. References provide fairly complex list structures. Each entry basically contains four main elements: the author(s)' name(s); the date of publication: the title of the work; and where it can be found. (This last element can be sub-divided m several ways,) There are numerous (conflicting) conventions f~,r presenting references. It is usual, for example, in an A1nericm~ journal to posiuon the date of publication after the journal title, thus: Dennis, 1. !the design and experimental testing of a hospital drug labelling system. Programmed Learnbg and k'ducationa! Technology, 1975, 12, 2, 88 94. whereas in a British !ournaa it would be more common to find the date of publication after the attthor's name, thus: Dennis, I. (l 975) The design and experimental testing of a hospital d rug labelling system. Programmed Learning a,M k'ducational Technology, 12, 2, 88 94. hi addilion to this major difference there are a number of mino~ ones. Some journals use capital letters for author(s)' ,qame(s), some do noL Some put double quotation marks around article titles, some single, some none at all. Some capltalise the first letter of each main word in the article title. Some use a full-stop after the title, some a comma. Mos| abbreviate journal tilles. So~'ne put a comma after tire journal titles, some a full-stop, some nothing. Some use bold typefaces for volume numbers, some italics, some exclude part numbers, and so on. In short, typographic cueing is used in a variety of ways to code the different elements of a reference entry, The possibility of using spatial cueing to achieve the same ends has not been widely practised, probably because it is thought that spatial cues will take up too much space. One journal, however, does use spatial cueing more than most. In Applied Ergonomics the same reference would appear thus: Dennis, 1. 1975 Programmed Learning and Educational Technology, 12.2, 88 94. The design and experimental testing of a hospital drug labelling systenr. In our view a more preferable layout would be: Dennis I. (1975) The design and experimental testing of a hospital drug labelling system. Programmed Learning and Educational Technology, 12, 2, 88 94. because this style presents the material in a more natural sequence and it separates the elements more clearly by placing each of the three main ones on a new line. The roles that can be played by spatial and typographic cues in references are of particular interest to the authors of this paper. Elsewhere, in previous articles, we have firmly advocated manipulating spatial cues rather than typographic ones to convey the structure of continuous text. We have argued that decisions about the spatial arrangements of the text on a particular page-size must come first : after this one might add in a typographic cue to enhance the spatial arrangement (see Hartley, 1978a, 1978b ). Sonre support for this position applied to the presentation of lists was found in a recent study on |he design of contents pages for social science journals (Hartley. 1979). In the present paper these ideas are exanlined in the context of designing references for articles in social science journals. The paper reports on tire preferences of experienced 0003 6870/79/03 165 -05 $02.00 @ IPC Business Press Applied Ergonomics September 1979 165

The role of spatial and typographic cues in the layout of journal references

Embed Size (px)

Citation preview

Page 1: The role of spatial and typographic cues in the layout of journal references

Applied Ergonomics 1979, 10.3, 165-169

The role of spatial and typographic cues in the layout ot: journal references James Hartley*, Mark Trueman* and Peter Burnhill t

*Department of Psychology, University of Keele. tDesign Department, Stafford College of Further Education.

Fifteen American and 15 British readers used the method of paired comparisons to assess 12 different ways of presenting reference lists in social science journals. The results indicated -- despite British preferences for British styles - that the spatial arrangements of elements in the list was the major determinant of preference, and that the presence of a typographic cue only enhanced a preference for a particular layout.

Many "kinds of text contain lists of items, each one of which mighl contain a number of sub-elements. Some lists are simple, such as contents pages in books, and some are more complex, such as lists of references or bibliographies. !n deciding how best to present such materials typographic designers have two related problems: first, how to make lists easy ,'o use, and second how, at the same time, to present material economically.

Such problems are usually overcome by the judicious use off typographic cueing (ie, the use of capitals, bold-faces, italics, etc). There has been little research, however, into the effectiveness of different ways of presenting the same list and virtually no research at all into the cost-effectiveness of, or the preferences for, different resolutions of the same problem. These latter issues are examined in this paper with reference to designing reference lists for articles in social science journals.

References provide fairly complex list structures. Each entry basically contains four main elements: the author(s) ' name(s); the date of publication: the title of the work; and where it can be found. (This last element can be sub-divided m several ways,) There are numerous (conflicting) conventions f~,r presenting references. It is usual, for example, in an A1nericm~ journal to posiuon the date of publication after the journal title, thus:

Dennis, 1. !the design and experimental testing of a hospital drug labelling system. Programmed Learnbg and k'ducationa! Technology, 1975, 12, 2, 88 94.

whereas in a British !ournaa it would be more common to find the date of publication after the attthor's name, thus:

Dennis, I. ( l 975) The design and experimental testing of a hospital d rug labelling system. Programmed Learning a,M k'ducational Technology, 12, 2, 88 94.

hi addilion to this major difference there are a number of mino~ ones. Some journals use capital letters for author(s) ' ,qame(s), some do noL Some put double quotation marks around article titles, some single, some none at all. Some capltalise the first letter of each main word in the article title. Some use a full-stop after the title, some a comma. Mos| abbreviate journal tilles. So~'ne put a comma after

tire journal titles, some a full-stop, some nothing. Some use bold typefaces for volume numbers, some italics, some exclude part numbers, and so on.

In short, typographic cueing is used in a variety of ways to code the different elements of a reference entry, The possibility of using spatial cueing to achieve the same ends has not been widely practised, probably because it is thought that spatial cues will take up too much space.

One journal, however, does use spatial cueing more than most. In Applied Ergonomics the same reference would appear thus:

Dennis, 1. 1975 Programmed Learning and Educational Technology,

12.2, 88 94. The design and experimental testing of a hospital drug labelling systenr.

In our view a more preferable layout would be:

Dennis I. (1975) The design and experimental testing of a hospital drug labelling system. Programmed Learning and Educational Technology, 12, 2, 88 94.

because this style presents the material in a more natural sequence and it separates the elements more clearly by placing each of the three main ones on a new line.

The roles that can be played by spatial and typographic cues in references are of particular interest to the authors of this paper. Elsewhere, in previous articles, we have firmly advocated manipulating spatial cues rather than typographic ones to convey the structure of continuous text. We have argued that decisions about the spatial arrangements of the text on a particular page-size must come first : after this one might add in a typographic cue to enhance the spatial arrangement (see Hartley, 1978a, 1978b ).

Sonre support for this position applied to the presentation of lists was found in a recent study on |he design of contents pages for social science journals (Hartley. 1979). In the present paper these ideas are exanlined in the context of designing references for articles in social science journals. The paper reports on tire preferences of experienced

0003 6870/79/03 165 -05 $02.00 @ IPC Business Press Applied Ergonomics September 1979 165

Page 2: The role of spatial and typographic cues in the layout of journal references

journal readers for different styles of references: styles Fig. l uses one item from the list to illustrate the six (a) which varied systematically in terms of spatial and typographic and the six (b) layouts.

cueing. A brief descriptive comment on each layout is as fbllows:

Designing different layouts

To test readers' preferences in this respect six layouts for a reference list were typed out with book and/or journal titles (and volume numbers) coded with an italic typeface, and six additional versions were typed out with the same spatial arrangement but without the italic cue. These two versions are labelled (a) and (b) respectively in this report.

Layout 1. This is a typical 'run-on ~ entry in the British style ie, there is no spatial separation of the main elements and the date of publication is given after the author(s)' name(s),

Layout 2. This is a typical 'run-on ' entry in the American style: ie, there is no spatial separation of the elements and the date of publication appears after the journal title.

la DiVesta, F. J. & Gray G. S. (1973) Listening and notetaking. Journal of Educational Psychology, 64, 278-287.

ib DiVesta, F. J. & Gray G. S. (1973) Listening and notetaking. Journal of Educational Psychology, 6h, 278-287.

2a

2b

DiVes~a, F. J. & Gray G. S., Listening and notetaking. Journal of Educational Psychology, 1973, 64, 278-287.

DiVesta, F. J. & Gray G. S., Listening and notetaking. Journal of Educational Psychology, 1973, 6h, 278-287.

3a

3b

DiVesta, F. J. & Gray, G. S. (1973) Listening and notetaking. Journal of Educational Psychology, 1973, 64, 278-287.

DiVesta, F. J. & Gray, G. S. (1973) Listening and notetaking. Journal of Educational Psychology, 1973, 6h, 278-287.

ha

hb

DiVesta, F. J. & Gray G. S. (1973) Listening and notetaking Journal of Educational Psychology, 64, 278-287.

DiVesta, F. J. & Gray G. S. (1973) Listening and notetaking Journal of Educational Psychology, 6h, 278-287.

5a

5b

DiVesta, F. J. & Gray, G. S. Listening and notetaking Journal of Educational Psychologg, 1973, 64, 278-287

DiVesta, F. J. & Gray, G. S. Listening and notetaking Journal of Educational Psychology, 1973, 64, 278-287

Fig. 1

6a DiVesta, F. J. & Gray, G. S. (1973) Listening and notetaking Journal of Educational Psychology, 1973, 64, 278-287.

6b DiVesta, F. J. & Gray, G. S. (1973) Listening and notetaking Journal of Educational Psycholo~-, 1973, 6h, 278-287.

An example of each of the different layouts used in the experiment

166 Applied Ergonomics September 1979

Page 3: The role of spatial and typographic cues in the layout of journal references

Table 1: The medians and preference rankings from American and British judges for references in the six styles -- (a) with and (b) without italic cues

US Judges UK Judges (N=15) (N=15)

Layout Median Range Median Range

UK style la 5 (0-9) 5 (0-11)

lb 2 (0-4) 4 (1--7)

US style 2a 4 (1-11 ) 3 (1-9)

2b 1 (0-6) 0 (0-4)

Compromise 3a 5 (1-9) 5 (1 - 10)

style 3b 3 (0-9) 2 (0-5)

UK style 4a 9 (3-11) 11 (6---11)

4b 6 (0-8) 8 (3 -10)

US style 5a 10 (3-11) 7 (3 10)

5b 7 (1-9) 3 (0 -9)

Compromise 6a 11 (4-11) 9 (5-11)

style 6b 8 (1-10) 6 (0-10)

Layout 3. This is a compromise solution. Tile date of publication is given both after the author(s)' name(s) and the journal title.

Layout 4. This is a "vertically-spaced' entry after the British style: ie, each main element name and date, title, and place of publication begins on a new line.

Layout 5. This is a 'vertically-spaced' entry in the American style: again each element starts on a new line.

Layout 6. This is a compromise 'vertically-spaced' entry.

Problems arise with these two main systems when authors have two or more publications in the same year. The British system solves this by using the device of(1972a) (1972b) etc, whereas the American one places the letters a or b in brackets after the end of the reference. The compromise solution allows one to put (1972a) after the author(s)' name(s) and to put 1972 in the element concerned with the place of publication. (See the references at the end of this article).

Another problem arises with lengthy references for example, a reference to an article in an edited set of readings. In this case, in this particular enquiry, spatial cues were not used to separate out further these elements. Thus such a reference would be presented (in the compromise style) in the following way:

Bumhill, P.. Hartley, J., Young, M. & Fraser, S. (1975) The typography of college prospectuses. In L. Evans and J. Leedham (Eds) Aspects of Educational Technology. London: Kogan Page, 1975.

It is important to note that in creating these sets of references we included the titles of the articles (some journals omit them) and we did not abbreviate the journal titles. In our view such abbreviations are not only cumbersome, but they also provide difficulties for students new to the discipline and for readers from other areas of expertise.

Each entry in the complete list was typed single spaced,

with a line space between entries for all 12 layouts. The length of tile lists (in terms of units of line-teed) was kept approximately constant: this meant that the run-on lists contained more entries than the vertically-spaced lists (8:6) as the latter took more units of line-teed.

The aim of creating these two sets of layouts was to assess (by using reader preferences) the relative importance of the roles played by spatial and typographic cues in the design of references.

Procedure

Fifteen American and 15 British participants took part in this study. The participants were either experienced university teachers and researchers, full-time research workers, or graduate research students in psychology: thus they were all familiar with using references. There were 11 men and four women in each group of 15.

Each participant was asked to judge the layouts in terms of how well-designed they considered them to be. It was pointed out to the judges, before they began the task, that a reference list could serve several functions. Three functions were especially mentioned, these were: (i) to refer to whilst reading the text to see if a paper

was readily accessible, or seemed to have an interesting title, etc;

(it) to refer to whilst searching library shelves in order to find a specific item in the list:

(iii) to dictate from or note from in an abbreviated form for some other user, or for oneself at a later date.

Each participant was then asked to judge each layout using the method of paired comparisons. The British judges compared the layouts in the order shown in Fig. 1 (ie, l a, l b, 2a, 2b, etc) but the American judges used the order (2a, 2b, la, lb, 3a, 3b, 5a, 5b, 4a, 4b, 6a, 6b). In other words, the judges started with the reference style that they (presumably) were most familiar with, although they did not necessarily label it 'British style', 'American style', or 'Compromise' . (Most participants in fact, seemed unaware of the different national conventions.)

Results

The results obtained from this procedure are shown in Table 1. Medians are presented for each layout in each version.

Sign tests carried out on the data shown in Table 1 indicated that the American judges did not significantly prefer the American style to the British one, and similarly, that they were not hostile to the compromise solution (ie, there were no significant differences between the preference rankings for layouts la, 2a, 3a and I b, 2b, 3b, and for layouts 4a, 5a, 6a and 4b, 5b, 6b). The results did, however, indicate that these judges significantly preferred the vertical layouts to the run-on ones (p < 0-001 ) and that they significantly preferred the (a) versions (those with the italic cues) to the (b) versions (those without them) (p < 0.001 ).

Similarly sign tests on the data for the British judges indicated that they too preferred the vertical layouts to their run-on counterparts (p < 0-00l ) and that they too preferred the (a) versions (those with the italic cue) to the (b) versions (those without them) (p < 0-001 ). However, the British results were markedly affected by a strong preference for the British style: thus a run-on layout in the British style

Applied Ergonomics September 1979 167

Page 4: The role of spatial and typographic cues in the layout of journal references

without italics was as acceptable as a vertical layout without italics in the American style (eg, compare layout 2b with layout 4b for the US and the UK participants). Because of these differences in national preferences it was not considered legitimate to pool the results shown in Table 1.

The similarities and differences in the results for the American and British participants were shown more clearly when we examined how many out of the 15 particpants judged in a particular way. The answers to a series of questions in this respect were as follows:

1. How many judges out of fifteen preferred the vertical layout to the run-on ones?

US Judges UK Judges 13 favoured vertical 14 favoured vertical 1 favoured run-on 1 favoured run-on 1 no clear preference

2. How many judges out of fifteen preferred the italic to the non-italic layouts.

US Judges 12 favoured italic 1 favoured italic in vertical, but not in run-on 2 no clear preference

UK Judges 14 favoured italic 1 favoured italic in vertical, but not in run-on

3a. How many judges preferred the US style to the UK style (run-on layouts)?

US Judges UK Judges 5 favoured US style 0 favoured US style 8 favoured UK style 13 favoured UK style 2 no clear preference 2 no clear preference

3b. How many judges preferred the US style to the UK style (vertical layouts)?

US Judges UK Judges 6 favoured US style 1 favoured US style 8 favoured UK style 14 favoured UK style 1 no clear preference

4a. How many judges preferred the compromise solution to the US/UK styles (run-on layouts with italic cues)

Table 2: The effects of line-length on the number of extra lines required for a vertically-spaced reference system

Column No of No of Average width lines for lines for number

six run-on six of extra references vertically- lines per

spaced reference references

Brit J. Psycho/ 135 mm (Pre 1977)

J. Appl Psycho/ 67 mm

'Comment' section of 50 mm American Psychologist

10 20 2-0

20 26 1-3

23 26 1-1

US Judges UK Judges thvoured compromise 3 favoured compromise

6 disliked compromise 5 disliked compromise 3 no clear preference '7 no clear preference

4 b How many judges preferred the compromise solution to the US/UK styles (vertical layouts with italic cues)

US" Judges UK Judges 7 favoured compromise 3 favoured compromise 4 disliked compromise 4 disliked compromise 4 no clear preference ~ no clear preference

With reference to questions 4a and 4b the category "no clear preference' indicates that the judges put the compromise solution in the middle - le. an order of preference might be UK. Compromise. US style.

Finally in this section, we may note here in addition that we asked the British judges to take their most preferred run-on layouts with the line spaces between entries and to compare them with their most preferred vertical layouts without the inter entry line spaces. In this situation 12 oul of the 15 judges still preferred tile vertical layouts

Discussion

The answers to the questions posed above indicate that both groups were similar in their preferences for vertically- spaced layovts combined with an italic cue, but that as noted above, there were differences between the preferences of the two groups for the different styles of presentation. Some American participants seemed to see value in the British system, and thus the compromise one (which incorporated elements of both solutions), whereas most of the British participants preferred the British style throughout.

Nonetheless. despite these findings, the results obtained in this enquiry support the argument put forward earlier in this paper, and they complement the results of Hartley (1979). For both groups the spatial arrangement of the list was the major determinant of the preference ranking, and the presence of a typographic cue enhanced the preference for a particular spatial layout. The average difference in the median rankings for the vertical over the run-on layouts was 5-I for the US judges and 4-2 for the British. The average difference in the median rankings for the italic over the non- italic layouts was 3-1 for both sets of judges.

The results of this inquiry must, of course, remain specific to the styles of reference list used and to the judges we employed. Some possibilities that were not considered were the use of capital letters for author(s)' names, quotation marks around the titles of articles, the abbreviation of journal titles, the use of bold typefaces tbr authors or Jbt journal volume numbers, the numbering of references (in the rex1 and list) and so on. Undoubtedly there will be adherents fi)r each of these variations (eg, see O'Donoghue. 1968. and Garfield. 1968).

Another issue which we did no1 consider was the tune taken ¢o retrieve items from reference lists printed with and without typographic cueing. Previous research has indicated that sometimes there is a correlation between preference and speed of retrieval (Tinker and Paterson. 1942) and sometimes not (Hartley etal. 1973). The little research that has been done on retrieving from references with and without typographic cueing has shown the time differen cos to be very small (see Hartley, 1978a). In our view such speed of retrieval measures are useful in so far as they provide

168 Applied Ergonomics September 1979

Page 5: The role of spatial and typographic cues in the layout of journal references

another measure of the clarity of layouts, but in this particular context we feel that the value of such data is partly academic. In practical terms one does not often indulge in a race against time when using a reference list.

A more relevant problem at this point is that of economics. Most readers will claim that the preferred solutions are expensive to produce because they take up more units of line-feed. (Indeed, this consideration led one or two judges to prefer the run-on layouts.) This claim is true, but we would argue that costs in this respect are not likely to be as dramatic as one might suppose. Most journals, for example, do not start new articles immediately following the references to the previous one : new articles typically start on a new page. Hence there is some room for manoeuvre here.

In this paper we shall examine two aspects of this problem: these are (i) the effects of column widths, and (it) the effects of extra lines on total costs.

The choice of column width is an important consideration in deciding whether or not to use a run-on or a vertically- spaced reference list. To examine the effects of presenting vertically-spaced reference lists with different line-lengths we typed out six references with matching line-lengths from each of three journals. The results obtained are shown in Table 2. They indicate that it is much more costly to use a vertically-spaced reference system in a wide column. (Incidentally, we should note here that it is of course possible to use a smaller typesize with the same interline space for the setting of references. We have illustrated elsewhere how this strategy might, in narrow column widths, produce text that is easier to read: see Hartley and Burnhill, 1978).

To calculate how many extra pages a change from a run-on to a vertically-spaced reference list might produce we examined the 1977 volumes of two journals: the British Journal of Psychology (4 issues) and the American Psycholc;gist (12 issues). The results indicated that as far as the British Journal was concerned it would appear that each issue would require on average two extra pages to accommodate a vertical reference system. However, as each issue already contains on average five blank left-hand pages (as new articles begin on right-hand pages) then - with some minor re-shuftling of the articles -- no more extra paper would be needed.

As far as the American Psychologist is concerned, it would appear that each issue would require approximately four extra pages to accommodate a vertical reference system. In this journal, however, new articles start on either the left or the right-hand pages - so there is little room for manoeuvre here. Four extra pages per issue, multiplied by the number of issues (12) and the circulation (approximately 52,000) leads to a considerable quantity. It may be, there- fore, that a vertical reference system may be unacceptable on economic grounds for journals which start articles on either left or right-hand pages, and which have a large circulation.

Conclusions

What then can we conclude? What guidelines for action can we give? It seems clear that readers prefer a vertically- spaced system preferably with an interline space between each entry -- but that this system is less economical in terms of space than is the traditional run-on style. It may well be possible, however, for some journals to use such a system

or to change to such a system without increasing costs. We would recommend that where this is possible, then it should be considered. Furthermore, where only small increases in costs might be expected then we would recommend that it might be worth considering how such costs could be offset (eg, by using smaller typesizes, or by increasing advertising revenue).

We also think that there is value in utilising the compromise solution to the American and British ways of presenting references. This system, as noted above, incorporates useful features from both styles. Most judges liked the reference list to match the text presentation with the author(s)' name(s) followed immediately by the date of publication. Accordingly, therefore, we have used the vertically spaced compromise system in presenting the references to this paper. Readers may now like to consider for themselves the advantages 'and limitations.

Acknowledgements

The American data cited in this article were collected when the senior author was a visiting member o f the Acoustical and Behavioural Research Center, Bell Laboratories, Murray Hill, New Jersey, USA. We are indebted to colleagues at Bell Laboratories and the University of Keele for acting as participants in this study. We are also indebted to the Social Science Research Council, UK, who funded this research.

References

Garfield, E (1968) Numerical versus alphabetic order for cited references. New Scientist, 1968, 39, 565 566.

Hartley, J. (1978a) Designing Instructional Text. London: Kogan Page. New York: Nichols.

Hartley, J. (1978b) Space and structure in instructional text. Paper to NATO conference on the Visual Processing of Information, Het Vennenbos, September. (Copies available from the author).

Hartley, J. (1979) Designing journal contents pages: the role of spatial and typographic cues. Manuscript submitted for publication. (Copies available from the author).

Hartley, J. & Burnhill, P. (1978) Fifty guidelines for improving instructional text. In J. Hartley and I.K. Davies (Eds) Contributions to an Educational Technology, Vol. 2 London: Kogan Page. New York: Nichols, 1978.

Hartley, J., Timson, S. & Burnhill, P. (1973) Subjective preference and the retrieval of information from reference materials. Visible Language, 1973, VII, 167~ 170.

O'Donoghue, P.N. (1968) Citations. New Scientist, 1968, 37, 321.

Tinker, M.A. & Paterson, D.G. (1942) Reader preferences and typography. Journal of Applied Psychology, 1942, 26, 38- 40.

Applied Ergonomics September 1979 169