9. Report the Results - Descriptive & Correlative

Introduction

Having analysed the data, you need to report the data.

A summary of the data and any analysis performed should be presented in the results section. The interpretation of, and conclusions from, the data are presented in the Discussion section of the report. All the results collected must be presented, not just the ones favourable to the hypothesis. The Results should include "the facts, the whole facts and nothing but the facts". The aim is to provide the reader with a concise and intelligent summary of the data collected and any analysis performed on that data.

Layout of Results Section.

In reporting the results it will probably be useful to describe the results related to each hypothesis in turn. Remember, there are 3 types of data. Below are examples of how to report each type of data.

Reporting numbers

Use Standard International units (metric). Write numbers as figure when 10 or greater, when with a unit of measurement, when a statistic. Write numbers as words when less than 10, when starting a sentence. Do not use 0 in a decimal number if the number cannot be greater than 1 eg. probability or corrlaltion. When reporting statistics, note degrees of freedom, (sample size), value and probability of statistic. Whatever type of research conducted, there will be desscriptive data to report.

Ways of Reporting Descriptive Data

Descriptive data is reported in 3 ways; text, tables & figures

Reporting Descriptive Data in Text

As with the text throughout the report, text reporting descriptive data should be concise and informative. Be careful to be explicit; do not make assumptions that the reader will know what you are talking about. Rather, ensure each section can be easily and unambigiously understood, even if read in isolation

A common fault with Results sectiopns is that they are difficult to read because they appear like a series of unconnected statements. The text of the Results section should lead the reader along a path, observing the data and analysis as you go. Make each section comprehensible and flow from one sectio to the next.

Be careful not to confuse your reader with statistics. Present what analysis done and what the statistical result clearly. Presenting the statistical analysis results unclearly suggests you don't understand what you did.

You should mentione in teh text the main points about the data and the analysis, but do not need to list all the data. Larger volumes of data are better presented in tables.

Reporting Descriptive Data in Tables

Tables provide a concise way of presenting lists if related results. Frequency counts, percentages, rankings, related measures, related statistics can all be usefully presented in tables.

Refer to every table in the text. For example, "as shown in Table 3, the..." or "relationship of the mean scores (see Table 4)" Highlight the main points, don't duplicate the information. Do not say "in teh Table above". Label tables numerically with arabic numerals.

Every table should be understandable in isolation. the table title and other headings and text in the table should be informative and concise. The title appears above the table. Always explain abbreviations and identify measurements.

Think carefully about the layout and order of the data in teh table to make it clear as possible for the reader. Often the independent variable (IV) is located in left column, with dependent variables (DV) in teh columns to the right. Make sure tables are of consistent layout throughout the report.

Reporting Descriptive Data in Figures

To effectively communicate some results figures should be used. Figures include graphs, drawings, photographs.

To help the reader perceive a pattern in the data, graphs can be used. Similarly, comparisons can be effectively shown with graphs. Common type of graphs used include polygons, histograms adn pie graphs.

Line graphs should be used when teh data are connected. The scales on both axes should result in the line accurately representing the data, that is not hiding or magnifying any change. A double slash should be made across the axis if it does not sart at zero.

Bar graphs should be used for data that is not connected. Usually teh IV (eg. group) on teh x axis, and the DV on the y axis. To give the reader a quick visual appreciation of the effect size compareed to variability, data variability can be shown by a "whisker" above representing one stadard deviation (for metric data). Use easily distinguishable shades and patterns to distinguish between IVs.

Pis graphs should be used for showing proprtions of a whole. Use shade and patterns to differentiate between segmetns, with the darkest shade for the smallest segment. Use less than five segments.

Scatter graphs should be used to show associated or lack of association between variables. To describe results that are difficult or lengthy to describe verbally, like postural symmetry, drawing or phorgraphs may be used.

As with tables, figures shoudl eb presented with sufficient information to be understandable in isolation. The figure caption should be clear, inforamtive and concise. It appears below the figure. The legend should concisely explain the symbols. Axes should be clearly labelled. Each figure should be referrd to in the text, as with tables.

Reporting Correlation Data

Like descriptive data, correlation data is reported in 3 ways; text, tables adn figures. The same principles of clarity and conciseness apply, remembering you are leading the reader through the data like a guide ona scenic tour. When using figures, teh x, y plot or scattergram is teh figure to use. This allows the reader to get a quick impression of the strength of any relationship, and whether it appears to be a linear relationship. Confidence interval for teh regression line can also be shown on the scattergram.