Back MatterSource: Proceedings: Biological Sciences, Vol. 264, No. 1389 (Dec. 22, 1997)Published by: The Royal SocietyStable URL: http://www.jstor.org/stable/51127 .

Accessed: 05/05/2014 04:50

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

.

The Royal Society is collaborating with JSTOR to digitize, preserve and extend access to Proceedings:Biological Sciences.

http://www.jstor.org

This content downloaded from 62.122.77.88 on Mon, 5 May 2014 04:50:08 AMAll use subject to JSTOR Terms and Conditions

Editorial code for presentation of statistical analyses

INTRODUCTORY GENERAL COMMENTS

The proliferation of statistical methods and software increases the need for authors to state, unambiguously and informatively, the methods they have used. The reader of any paper should expect to learn precisely what the author has done, and to be able to adopt similar procedures in his or her own work.

An author who uses an algebraic symbol or technical term that is not common enough to be generally under- stood should state exactly (by definition or by bibliographic reference) the meaning given to it. Here we set out a code for the statistical terms and symbols that we regard as part of the general language of science. The Editor will accept them without detailed

explanations or definitions. There can be no absolute standards of correctness: new developments in statistical methods continually appear, and statisticians them- selves do not employ a unique system of symbols and nomenclature. Confusion is inevitable, however, if authors fail to make clear their own usages.

THE CODE

The code that follows is not intended to restrict an author's freedom to choose symbols or statistical methods, but only to simplify the task of writing. The Editor will accept the following without definition, but will expect any different usages to be made explicit: t: Student's t-statistic (when used for a difference

between two arithmetic means, the text must

clearly identify the means that are being compared).

r: Product moment or Pearson correlation coeffi- cient (not any other), modifiable to r,y if the variables involved need to be specified:

b: Simple linear regression coefficient ofy on x, where x,y have already been clearly identified, modifiable to by,x if the variables need to be

specified. The word 'slope' is not an acceptable synonym for 'regression coefficient'.

F: Ratio of two independent error mean squares, usually, but not invariably, in an analysis of variance (degrees of freedom must be stated).

x2: Statistic used in many well-known tests of significance relating to tables of frequencies, where, if correctly calculated, it has a prob- ability distribution approximately that of a sum of squares of several independent N(0,1) variables,'N(0,1 )' referring to a Normal distri- bution with mean 0 and variance 1. The degrees of freedom must be stated. If it is necessary to do so, this statistic can be described in the text as 'chi-squared', but it is

preferable to use the symbol. p: Probability obtained in the course of applying

a test of statistical significance (it should be

+: This sign can be used to introduce the standard error of the mean whose numerical value has preceded it. It is perhaps better avoided, because some have used it with very different meanings.

s.d.: Standard deviation, the square root of an esti- mate of variance.

s.e.: Standard error, always of a mean or of an esti- mate of some other parameter.

d.f.: Degrees of freedom.

log: Unless otherwise indicated by subscript, always the natural logarithm (base e).

loge: Natural logarithm (base e). log10: Logarithm to the base 10.

Using any of these abbreviations or symbols with different meanings (for example, using t for a time, or r for a ratio) requires that they be defined. However, the use of F, N or P in chemical contexts without defini- tion is unlikely to cause misunderstandings. All referees are asked to take note of any departures from usages as defined above, and to regard them as requiring that a paper be returned to its author for clarification. In particular, any use of N or n should include a statement of whether this is the number of observations, the number of degrees of freedom, or something else. Authors need not be inhibited from using variants of these symbols, such as 'r' for a Spearman rank correla- tion coefficient, provided that the symbolism is defined.

The description of X2 as 'well-known' does not imply that it can be used uncritically. Many of the commonest errors in the elementary application of statistical methods are associated with the misuse or miscalcula- tion of this statistic, but for comparing observed and expected frequencies it remains one of the most widely used techniques. Mention of F does not imply that its numerical value must always be given, for often no great interest attaches to it.

Every mention of t or For x2 should explicitly state the degrees of freedom, either as subscripts or in brackets, for example, as t17, F3 27, or F[3,27], 28], and so on, unless the accompanying'text makes the numbers of d.f. apparent. For a correlation coefficient, the degrees of freedom should always be stated unless the number is evident as being two less than a previously stated total number of pairs of values.

Authors should remember that the physical dimen- sions of any s.d. or s.e. are those of the measurement to which it relates. A regression coefficient ofy upon x has dimensions 'units of y per unit of x', and an explicit statement of these units may be needed in any discus- sion of a particular coefficient. The statistics t, r, Fand Z2 are dimensionless, as are some statistics that appear in less-familiar methods. For all others, the physical dimensions should be stated wherever numerical values are shown unless the text has made the units very obvious. Any unstated assumption that time is printed in the lower case).

This content downloaded from 62.122.77.88 on Mon, 5 May 2014 04:50:08 AMAll use subject to JSTOR Terms and Conditions

Statistics editorial

measured in days or in seconds, or that mass is measured in kilograms, may confuse a reader!

Within the written text of a paper, statistical symbols should be printed in italics rather than roman letters, except in circumstances where this practice might itself cause confusion. Special care is needed with the names of methods. Some, such as 'multivariate analysis' and 'cluster analysis', relate to broad classes of statistical

technique, and are likely to be ambiguous unless they are specified more exactly. Research workers often give newly fashioned names to new concepts, and whether or not these come into widespread use is largely deter- mined by natural selection. Some methods are

commonly named eponymously: convenient though this practice is, such a name may be meaningless even to a professional statistician who is not familiar with a

highly specialized field of application. In all cases of

doubt, the Editor will expect an author to make clear the meaning, and to provide appropriate bibliographic references for any method whose name has not become

generally known in the particular field of application. Many biologists today analyse their data by using

statistical packages on mainframe or desktop com-

puters. Among the ever-growing number of general- purpose packages available today, a few have become so widely used that their names can now be regarded as assimilated to the general language of science.

Hence, in the text of a paper, a reference to BMDP, GENSTAT, GLIM, SAS or SPSS is admissible without further identification. This is not intended in any way to restrict authors in their choice of software, or even as recommending particular statistical packages. How-

ever, an author should describe, in the section headed 'Materials and methods' or elsewhere, any use of a different package, exactly as the standard conventions of scientific writing would lead him or her to do in

respect of special apparatus or special reagents, with

appropriate identification of its source and the address of its producer or supplier. When using any software

package, an author cannot safely assume that the

computer output will be presented solely in language and with terminology that is totally acceptable without explanation. The single aim of this recommen- dation is to ensure that what an author has done is made clear to the readers.

Authors should seek to avoid giving a spurious appearance of numerical exactness. Few sets of bio-

logical data can sensibly be summarized by means, special-purpose coefficients, or other statistics pur- porting to be accurate to 1 part in 10 000. For example, it would be inappropriate to record all the numerals if the mean mass of an organism, given by a computer program, came out as 42.57146 mg. The same applies to probabilities obtained by the application of any significance test. Arithmetically correct computation might seem to lead to ~p= 0.0002583', but it is most unli-

kely that the underlying assumptions about the nature and source of the data would justify anything more exact than b= 0.0003', or possibly only p < 0.005'.

DavidJ. Finney John L. Harper December 1993

This content downloaded from 62.122.77.88 on Mon, 5 May 2014 04:50:08 AMAll use subject to JSTOR Terms and Conditions

Instructions to Authors (see also www.pubs.royalsoc.ac.uk)

GENERAL

Proceedings Series B welcomes papers on any aspect of

biological science. It publishes announcements of impor- tant results, from 1 January 1998 of no more than 4000 words maximum length, including a brief summary, references and typically four figures and/or tables; equivalent to no more than six printed pages. Authors are advised that papers prepared in accordance with these instructions will be given priority. Acceptance of a

paper will be determined by its quality and interest.

Papers longer than six printed pages will be consid-

ered, however, they will attract a page charge of C98

per printed page exceeding this limit. As an alternative to this, subject to the approval of the Editor, we are able to place electronic appendices onto the Royal Society Publications' Web Site (see Submitting Electronic Appendices below).

Papers describing experiments with vertebrate animals will be accepted only if the procedures used are clearly described and conform to the British Home Office regulations for avoiding unnecessary suffering to the animals.

SUBMISSION

Submitted papers must not have been published previously, nor be under consideration for publication elsewhere. Submission to Proceedings B requires that authors accept that the Editor has final authority in all matters relating to the acceptability and propriety of

publication. Authors should send papers to the Proceedings B

Editorial Office, The Royal Society, 6 Carlton House Terrace, London SW1Y 5AG, UK. Please indicate to whom correspondence and page proofs should be addressed, and include daytime telephone and fax

numbers, and e-mail addresses. The date of the paper's receipt will be published if the paper is accepted.

Four complete copies of the typescript (and any appendices) and of all figures (together with one

floppy disk for the text) should be sent. We require one

complete set of figures as original drawings or glossy prints, (i.e. suitable for reproduction). In the case of

photographs, the extra copies should preferably be

glossy prints, but photocopies for the use of referees are permissible provided that they show the features described in the paper. Before dispatch, label and date all typescripts, figure sets and the disk(s) so that they can be easily identified.

Once accepted and finalized, your paper is passed to our Journals Production Office. Any enquiries at this

stage should be addressed to them on + 44 (0) 171-839 5561.

FLOPPY DISKS

A floppy disk is required for production of your paper. We can use most PC or Macintosh word-processing and

text formats, including TEX files. We encourage the submission of figures on disk, as separate PostScript files.

COPY

Papers should be typed. All sheets should be numbered

serially and securely clipped together. Check the type- script carefully before submitting. Spelling should conform to the preferred spelling of the Shorter Oxford English Dictionary.

STRUCTURE

It is important that the structure of your paper is consistent with the rest of the journal. Please ensure that it contains the following in this order: a title; authors' names and place of work; a brief, comprehen- sive summary; the main text of the work; with any acknowledgements and references listed at the back; a list of figure captions, numbered in sequence as they appear in the paper; a short title for page headings, and six key index words or phrases. Papers should be divided into sections, described by short headings. Materials and methods sections should be marked in the margin for small type. Footnotes are not used.

TITLE AND SUMMARY

It is essential that the title and summary should be as short and as general as possible. The title should also

appear on a separate covering sheet which should also bear the authors' names and the addresses where the work described was carried out. The current addresses of authors, where different to these, should be indi- cated, as well as the author to whom correspondence should be sent.

To this sheet add a short title (maximum of 50 letters and spaces) suitable for page headings, and up to six

(specific and useful) keywords for inclusion in the

subject index. The summary should be comprehensive and intelli-

gible by itself, and must not contain references or

unexplained abbreviations. It must not exceed 200 words.

SUBMITTING ELECTRONIC APPENDICES

Subject to the approval of the Editor, we are able to

place electronic appendices (created by authors them-

selves) onto the Royal Society's Publications' Web Site. A hard copy will be deposited with the British Library. The main report, published in the printed journal, should be able to stand on its own merit.

Any electronic appendices will be refereed together with the main body of the paper. On acceptance, the final version of an appendix will appear on the Royal Society's Publications' Web Site. This material should

This content downloaded from 62.122.77.88 on Mon, 5 May 2014 04:50:08 AMAll use subject to JSTOR Terms and Conditions

be supplied to us in an electronic format ready to view. It is the responsibility of the author to check for spel- ling, grammar and presentation. The appendix should be supplied as a word-processing file or in HTML

(HyperText Mark-up Language). Please do not supply it as a PDF (Portable Document Format).

UNITS, SYMBOLS AND ABBREVIATIONS

If special characters are used, list these on a covering sheet (and make marginal notes to indicate their posi- tion), taking special care to spell out ambiguous characters, such as Greek letters.

As far as possible the recommendations based on the International System of Units (SI) contained in Quan- tities, units and symbols (1975, The Royal Society, ?3.50) should be followed.

Wherever possible, only internationally agreed abbreviations should be used; see, for example, the list of accepted abbreviations for use in the Biochemical Journal.

STATISTICS

As far as possible, the presentation of statistics should follow the guidelines published each year in the December issue of the Proceedings. When referring to computer programs, authors should

specify clearly the procedure used, and should quote publications that will allow the reader to ascertain how they are carried out.

ILLUSTRATIONS

All illustrations should fit within the text, into single column (84mm) or double column (175mm) widths. It is essential to get the line widths, shading and size of

labelling right, otherwise it may not be possible to use them. Figures are numbered in one sequence throughout the paper.

Labels should be added to the original drawings before submission if you are able to set lettering of suitable quality and size. We recommend the Times Roman font with a final reduced size of 9 point, or a

height of around 2 mm for capital letters. If in any doubt, please submit one set of unlabelled, original drawings in addition to the labelled drawings.

Line drawings. Descriptions should be placed when- ever possible in the legends and not on the figures themselves. Labels should be brief, e.g. (a), (b), etc., and explained in the legend. All lettering should be in lower case except at the start of sentences or in proper names. Keys should be drawn within the artwork area. Never use hairlines; although they appear okay on laser

output, they will be much finer when rendered at high resolution and may break up when printed.

Photographs. We need glossy prints, with the top edge indicated. The area covered must be limited to the

subject in question, or to a minmum representative area in photomicrographs, etc. This enables the photo- graph to be reproduced at the largest possible scale. Each micrograph must include a scale bar, applied to the original, with an indication of the exact length.

Captions should be typed on a separate sheet at the end of the paper. Use the following style:

Figure 7. Time-course of changes in fibre type composi- tion during post-stimulation recovery. (a) Type 1 fibres.

(b) Type 2A fibres, including the transitional fibres

(asterisks) referred to in the text. (c) Type 2B fibres. Bands indicate the range (mean + s.d.) for the corre-

sponding fibre type in control muscles.

The position of each illustration/table should be

clearly marked in the margin of the typescript. Identify each figure with the author's name and its number.

SUPPLYING FIGURES ON DISK

Please supply on a separate disk, ideally as PostScript, generated in a technical drawing package, such as Adobe Illustrator. Alternatively, we can handle EPS or TIFF formats provided that any scanning is done with the final output in mind, i.e. 1200 dots per inch (dpi) for line art and 300dpi for grayscale figures. Colour must be in CMYK mode. All figures should be

prepared with the final size, after reduction, in mind. Labels should be legible and conform to the journal's style. Please supply a hard copy in any case.

COLOUR

Proceedings B welcomes the submission of papers containing colour illustrations. However, owing to the

high cost of colour reproduction, authors are asked to contribute towards the extra cost. An estimate will be sent to you on submission. Exemption from charges for colour illustrations will only be made if referees decide colour is essential, and if authors make an explicit case to the Editor explaining why they are unable to obtain the necessary funds.

TABLES

Tables, however small, should be numbered and referred to in the text by their numbers (e.g. see table

3). Table headings should be brief; with descriptions of

experimental detail directly beneath, in parentheses. Column headings should be in lower case lettering, except for initial capitals and proper names. The units of measurement and any numerical factors should be

placed unambiguously at the head of the column, e.g. mass (g), yield (t ha-1), pmol ml-1 plasma.

REFERENCES

All references to the literature cited must be given in

alphabetical order at the end of the paper, and should be arranged as follows.

1. Author surname(s) with initials 2. Year of publication. 3. Title of paper or book. 4. Title of the periodical, abbreviated where possible, in

italic. 5. Volume number in bold. 6. First and last page numbers (separated by an en

dash).

This content downloaded from 62.122.77.88 on Mon, 5 May 2014 04:50:08 AMAll use subject to JSTOR Terms and Conditions

7. For a book, the edition, the chapter(s), the page ranges, the place of publication (if it is not obvious) and the name of the publisher should be given: e.g.

Andersson, M. 1994 Sexual selection. Princeton University Press.

Darwin, C. 1859 On the origin of species by means of natural selec- tion. London: John Murray.

Falconer, D. S. 1981 Introduction to quantitative genetics. 2nd edn. London: Longman.

Falkenmark, M. 1993 Landscape as life support provider: water- related limitations. In Population-the complex reality (ed. F. Graham-Smith), pp. 103116. London: The Royal Society.

Nilsson, L. A. 1988 The evolution of flowers with deep corolla tubes. Nature, Lond. 334, 147149.

References in the text are listed alphabetically by giving the names of the authors and the date of publication: e.g.

It has been described frequently elsewhere by Brown et al. (1972), Brown (1974), Clarke (1974), Clarke &

King (1974) and Brown & Jones (1982).

References by number are not permitted.

PAGE PROOFS

Because of the need for fast publication, only a few days may be available for checking proofs, so authors who

may be absent from their normal address must either inform the Society of their intended whereabouts or make arrangements for the proofs to be checked

quickly. Proof corrections may be returned by fax (on

+44 (0)171 976 1837). Authors are liable for the cost of substantial alterations to their proofs.

Fifty offprints of each paper are supplied free of

charge; further copies may be ordered at extra cost at

proof stage.

COPYRIGHT

In order to give the Royal Society authority to deal with matters of copyright, each author will be asked to

assign to the Society the copyright in any article

published in the journal. In assigning copyright, authors will not be forfeiting the right to use their

original material elsewhere subsequently. This may be done without seeking permission and subject only to normal acknowledgements to the journal. However, it would be appreciated if authors would inform the

Society in this event.

Authorship should be restricted to those who have contributed substantially to the work. The corre-

sponding author must ensure that all authors assign copyright for the paper. All authors should ensure that all necessary permissions have been obtained for repro- ducing any other copyright material including permission to reproduce and publish electronically anywhere in the world.

[December 1997]

This content downloaded from 62.122.77.88 on Mon, 5 May 2014 04:50:08 AMAll use subject to JSTOR Terms and Conditions

FUTURE PAPERS IN PROCEEDINGS SERIES B

Proceedings: series B publishes original papers in all aspects of the biological sciences, including those of an interdisciplinary or multidisciplinary nature. Papers up to 4000 words long are welcomed, particularly announcements of important new developments in biology. Reviews containing original and interesting ideas, and criticisms of papers already published, are also invited. Papers will be published rapidly (the target time being 3 months from receipt).

Future issues of the journal will include the papers listed below.

B. Furlow, S. W. Gangestad & T. Armijo-Prewitt Developmental stability and human violence

D. J. D. Earn, P. Rohani & B. T. Grenfell Persistence, chaos and synchrony in ecology and epidemiology

A. E. Cohen, A. Gonzalez, J. H. Lawton, O. L. Petchey, D. Wildman & J. E. Cohen A novel experimental apparatus to study the impact of white noiseand l/f noise on animal populations

J. M. J. Travis & C. Dytham The evolution of dispersal in a metapopulation: a spatially explicit, individual-based model

N. J. Savill & P. Hogeweg Spatially induced speciation prevents extinction: the evolution of dispersal distance in oscillatory predator-prey models

M. Loreau Ecosystem development explained by competition within and between material cycles

P. W. de Jong & P. M. Brakefield Climate and change in dines for melanism in the two-spot ladybird, Adalia bipunctata (Coleoptera: Coccinellidae)

M. C. Anstett, M. Giberau & M. Hossaert-McKey Partial avoidance of female inflorescences of a dioecious fig by their mutualistic pollinating wasps

P. Heeb, I. Werner, M. K6lliker & H. Richner Benefits of induced host responses against an ectoparasite

J. Prenter, R. W. Elwood & W. I. Montgomery No association between sexual size dimorphism and life histories in spiders

G. S. Hollrigel, R. J. Morris & I. Soltesz Enhanced bursts of IPSCs in dentate granule cells in mice with regionally inhibited long-term potentiation

M. Mon-Williams, J. R. Tresilian, N. C. Strang, P. Kochhar & J. P. Wann Improving vision: neural compensation for optical defocus

Cover picture from the paper by Bateson & Bradshaw. A deer hunt, from George Turbevile's book The Noble Arte of Venerie or Hunting, originally published in 1576. (The woodcut reproduced here is taken from the second edition, published in 1611 by T. Purfoot, London, and is reprinted with permission of the Syndics of Cambridge University Library.)

This content downloaded from 62.122.77.88 on Mon, 5 May 2014 04:50:08 AMAll use subject to JSTOR Terms and Conditions

PROCEEDINGS: BIOLOGICAL SCIENCES Proceedings of the Royal Society of London

Series B Volume 264 Number 1389 22 December 1997

CONTENTS

pages

P. BATESON & E. L. BRADSHAW

Physiological effects of hunting red deer (Cervus elaphus) 1707-1714 J. F. PRINZ & P. W. LUCAS

An optimization model for mastication and swallowing in mammals 1715-1721 L. A. WILKENS, D. F. RUSSELL, X. PEI & C. GURGENS

The paddlefish rostrum functions as an electrosensory antenna in plankton feeding 1723-1729 M. J. L. MAGRATH & M. A. ELGAR

Paternal care declines with increased opportunity for extra-pair matings in fairy martins 1731-1736 P. C. L. WHITE, A. J. G. LEWIS & S. HARRIS

Fertility control as a means of controlling bovine tuberculosis in badger (Meles meles) populations in south-west England: predictions from a spatial stochastic simulation model 1737-1747

J. A. MCKENZIE Stress and asymmetry during arrested development of the Australian sheep blowfly 1749-1756

D. C. KRAKAUER & R. J. H. PAYNE The evolution of virus-induced apoptosis 1757-1762

T. PATARNELLO, L. BARGELLONI, E. BONCINELLI, F. SPADA, M. PANNESE & V. BROCCOLI

Evolution of Emx genes and brain development in vertebrates 1763-1766 R. SPRENGELMEYER, A. W. YOUNG, I. PUNDT, A. SPRENGELMEYER, A. J. CALDER, G. BERRIOS, R. WINKEL, W. VOLLMOELLER, W. KUHN, G. SARTORY & H. PRZUNTEK

Disgust implicated in obsessive-compulsive disorder 1767-1773 R. BADDELEY, L. F. ABBOTT, M. C. A. BOOTH, F. SENGPIEL, T. FREEMAN, E. A. WAKEMAN & E. T. ROLLS

Responses of neurons in primary and inferior temporal visual cortices to natural scenes 1775-1783 C. J. PRICE & K. J. FRISTON

The temporal dynamics of reading: a PET study 1785-1791 M. M. JH. VAN OPPEN, G. F. TURNER, C. RICO, J. C. DEUTSCH, K. M. IBRAHIM, R. L. ROBINSON & G. M. HEWITT

Unusually fine-scale genetic structuring found in rapidly speciating Malawi cichlid fishes 1793-1802 G. A. PARKER, M. A. BALL, P. STOCKLEY & M. J. G. GAGE

Sperm competition games: a prospective analysis of risk assessment 1803-1812 S. PATERSON & J. M. PEMBERTON

No evidence for major histocompatibility complex-dependent mating patterns in a free-living ruminant population 1813-1819

K. J. O'CALLAGHAN, M. R. DAVEY & E. C. COCKING

Xylem colonization of the legume Sesbania rostrata by Azorhizobium caulinodans 1821-1826 M. MOLLER & Q. C. B. CRONK

Phylogeny and disjunct distribution: evolution of Saintpaulia (Gesneriaceae) 1827-1836 A. HASTINGS, S. HARRISON & K. MCCANN

Unexpected spatial patterns in an insect outbreak match a predator diffusion model 1837-1840 0. L. PETCHEY, A. GONZALEZ & H. B. WILSON

Effects on population persistence: the interaction between environmental noise colour, intraspecific competition and space 1841-1847

ERRATA 1849-1850 INDEXES 1851-1865

VOLUME TITLE PAGE

VOLUME TABLE OF CONTENTS

EDITORIAL CODE FOR PRESENTATION OF STATISTICAL ANALYSES

INSTRUCTIONS TO AUTHORS

Published in Great Britain by the Royal Society, 6 Carlton House Terrace, London SWlY 5AG1111111111111 111111111 1 111 Printed in Great Britain for the Royal Society 0962-8452(199712)264:1 389 by the University Press, Cambridge

This content downloaded from 62.122.77.88 on Mon, 5 May 2014 04:50:08 AMAll use subject to JSTOR Terms and Conditions