The assiduous reader of the previous chapters will already have gathered many notes about the future(s) of the portal. To pick out a few points: Franklin specified some next steps within the possibilities of the emerging technology. Winship reviewed the vast array of functions and services appearing in web portals, all candidates to be the killer application that sells a new organizational portal to users. Hepworth, Probets, Qutaishat and Walton pondered the possibilities wrapped up in the concept of personalization, concluding with a positive view of the forces gathering behind it. Sugianto and Tojib argued that a major trend will be towards facilitating access to the portal from anywhere with any device. Several of the public sector chapters seem to suggest that a portal is nearly always something that is going to happen, as much as something available now.

The chapters in this last section of the volume pick up these issues and look exclusively at future trends. The first is from the perspective of a dynamic force in the tertiary education sector, JISC. Notay takes a look at cross-sectoral developments which will probably create the infrastructure on which diverse web portals and local portals will be built. Bryson examines the potential role of portals for the arts and humanities, in the context of the Grid and e-Research. In the concluding chapter, Awre explores the possibilities held in the buzzworld of Web 2.0.

In the rest of this section introduction we consider another direction of development. Firestone and others have argued that the Enterprise Information Portal will evolve into a ‘knowledge portal’ (e.g. Firestone, 2003). What precisely that means is unclear, though Firestone makes his own suggestions. One thing that knowledge management has contributed to our view of the world is to reinforce our understanding of the diversity of information/knowledge that people need in the workplace – beyond books and data, even internal reports. It may be important and possible to in some sense manage knowledge in people's heads and knowledge that is bound up in communities. We have already seen in Musgrave's chapter how the development of portals needs to take into account pre-existing communities and cultures. This concept focuses more on the portal as a host and creator of communities.

This is the point everything has been leading up to. Having carried out the research and marshalled all the evidence, you are now faced with the problem of making sense of it all. Here you need to distinguish clearly between three different things: results, conclusions and recommendations.

Results are what you have found through the research. They are more than just the raw data that you have collected. They are the processed findings of the work – what you have been analysing and striving to understand. In total, the results form the picture that you have uncovered through your research. Results are neutral. They clearly depend on the nature of the questions asked but, given a particular set of questions, the results should not be contentious – there should be no debate about whether or not 63 per cent of respondents said ‘yes’ to question 16.

When you consider the results you can draw conclusions based on them. These are less neutral as you are putting your interpretation on the results and thus introducing a degree of subjectivity. Some research is simply descriptive – the final report merely presents the results. In most cases, though, you will want to interpret them, saying what they mean for you – drawing conclusions.

These conclusions might arise from a comparison between your results and the findings of other studies. They will, almost certainly, be developed with reference to the aim and objectives of the research. While there will be no debate over the results, the conclusions could well be contentious. Someone else might interpret the results differently, arriving at different conclusions. For this reason you need to support your conclusions with structured, logical reasoning.

Having drawn your conclusions you can then make recommendations. These should flow from your conclusions. They are suggestions about action that might be taken by people or organizations in the light of the conclusions that you have drawn from the results of the research. Like the conclusions, the recommendations may be open to debate. You may feel that, on the basis of your conclusions, the organization you have been studying should do this, that or the other.

Methods are the tools of the researcher's trade. You need to know how to use them but, just as important, is knowing when they should be used. The perfect researcher will be familiar with the widest possible range of methods and will deploy them selectively to meet the requirements of different circumstances.

For the rest of us, our aspirations are more modest. Most of us can only hope to be familiar with a limited range of methods, understanding their strengths and weaknesses and being aware of when to use one in preference to another. With luck, we will also know that there are lots of other different ways of approaching the collection and analysis of data and we will be in a position to identify when we need to extend our personal range of skills by learning to use a new method.

This part of the book aims to give you an overview of the range of social research methods that are available to you. It will indicate the strengths and weaknesses of each method and will suggest when and under which circumstances, they should be used. It will also give you some hints and tips that will help you get going. What it will not do is provide a detailed treatise on social research methods in all their sophisticated glory. There are plenty of other books that do this. Some of them have even been written by people who have actually used the methods. A selection of them is listed in the guide to further reading in Chapter 16.

At this point it is worth saying a word about triangulation. Many researchers use only one method at a time to collect data – the research project will consist of a questionnaire survey, or a series of focus groups. There is, however, much to be gained by using more than one method to explore an issue from different perspectives. A relatively large scale questionnaire survey will, for example, provide a broad picture. This picture can be enriched by a series of focus groups to explore aspects of the topic in more depth. The whole thing can be put into a wider context by a thorough review of the literature and consideration of the results of previous research on the topic. The jargon term for looking at things in this way is triangulation.

Just as a builder requires a detailed set of plans to guide the building of a house, so a researcher needs a proposal to help structure and manage the research project. Too often, the proposal is regarded as an inconvenience that is only required in order to obtain external funds for a project or to obtain approval for an academic dissertation. It is thought to be something that can be discarded once the funds have been allocated or the approval given. This is a very limited view of research proposals. Those who take it leave themselves without one of the most useful tools for managing research.

Research proposals, in fact, play a number of important rôles in the research process and, while preparing a proposal may seem like a distraction or an unnecessary waste of time, the work involved should ensure that you think through all the aspects concerned with the project before the research itself begins to take over. Because once the project starts rolling it is often too late to begin making changes and adjustments that would have been obvious in advance if the project had been properly prepared.

The purposes of research proposals

Proposals have a number of different purposes. A small in-house project will clearly require a less complicated proposal than a major project for which external funds are being sought. But the purposes they serve will, by and large, be the same.

To gain the approval of a supervisor

The proposal plays an important rôle in justifying the research to a supervisor. It is the vehicle that you use to argue the case for the research, demonstrating that it is important that you are capable of undertaking – and successfully completing – the work, and that the work will enable you to achieve the educational objectives of the exercise.

For academic research projects, you will need to obtain your supervisor's approval for your ideas at an early stage. You will need to be able to show that you have a clear understanding of the general issues and theories associated with the topic that you wish to study. You should be able to cite the key authors in the field and you should be able to demonstrate how your approach fits with the research that others have carried out.

For many people, collecting quantitative data is the core of social research. Types of quantitative techniques are the things that first come to mind when people think of social research – techniques such as self-completion questionnaires and interview surveys. They are all founded on the principle that you can generate information about the characteristics of a group by collecting data from a subset of that group.

This principle was developed first by researchers in the natural sciences. They carried out experiments and measured the results. The experiments were then repeated many times until the researchers could be sure that the results they obtained were not the product of random chance but were characteristic of the subjects or objects they were studying. This approach has been subsequently adapted by social scientists and it underpins all quantitative methods.

Sampling

Let's face it, sampling is difficult. The basic principle is straightforward: you take a sample from a large group, look in detail at that sample and then infer the characteristics of the whole group from those of the sample. Of course, it is not quite that simple. You can never be sure that the sample has the same characteristics as the group – they probably do, but you cannot be certain. So you have to allow for that degree of probability.

Sample size

You need a sample that is big enough to represent all the characteristics of the larger group. In part, this depends on the relative size of the group and the sample you select. Let us suppose that you are trying to measure the proportion of people in the population who have brown hair and blue eyes. You walk down the street and the tenth person you meet has brown hair and blue eyes. Can you infer from this that ten per cent of the population have brown hair and blue eyes? Probably not.

So you walk on and look at 100 people. Of these, eight have brown hair and blue eyes. So, is the proportion eight per cent? Well possibly, but to be a little more certain you carry on walking until you have passed 1000 people. By now you have counted 78 people with brown hair and blue eyes. You can fairly confidently say that about eight per cent of the population have the brown hair–blue eye combination.

Quantitative research aims to show you what is happening. Qualitative research, on the other hand, sets out to tell you why it is happening. It is all about developing a detailed understanding of individuals’ views, attitudes and behaviour.

The approaches to collecting qualitative data are much less structured and formal than the techniques used for gathering quantitative data. The aim is to allow respondents to talk, often at great length, about their feelings, and about their underlying attitudes, beliefs and values.

Some qualitative research gathers data from individuals. Here, semistructured and depth interviews are used to explore issues outside the tight constraints of a structured interview survey. Other techniques use groups of people to obtain a more considered picture, building on discussion and the development of individual views within a group setting.

Semi-structured interviews

For most people, the first step they take towards collecting qualitative data is through the use of semi-structured questionnaires. These can be thought of as a half-way house between the rigid formality of a structured interview – where the researcher attempts to fix and control the circumstances of the interview so that the data are collected in as consistent a fashion as possible – and the flexibility and responsiveness of a depth interview.

They are best used when you want to collect both structured information and information about attitudes or beliefs. This objective tends to determine the size of the sample to be used. The need to process some of the data in a quantitative way suggests that the smallest sample you can realistically contemplate is around 100 people. On the other hand, the need to make sense of the unstructured data will limit the sample to a maximum of about 200 people – more than that and you will end up analysing the qualitative data in a quantitative manner, so negating the purpose of the exercise.

You design the interview schedule in much the same sort of way as you would a structured schedule. Many of the questions will be closed, offering the respondent a limited range of options with the questions being asked in a pre determined sequence. Other questions will be open-ended. The interviewer will be asked to record the answers verbatim, either writing the answer down or recording it on tape.

To get the full value from your research you must be prepared to disseminate the results. You need to communicate information about the research and its results to those people and organizations that can make use of it. The art of good dissemination lies in your ability to select the dissemination channels that best suit the subject of the research and the nature of the audience.

Dissemination does not begin when the report is produced – you can start the process right from the outset of the project, alerting people to the existence of the research and preparing them for the eventual release of the results.

When thinking about dissemination it helps to think of three different audiences. The first have a stake in the research itself. They are actively awaiting the results and will want to see the full report. This group includes the research funders and all the people and organizations that are expected to take some action as a result of the work.

It is not difficult to disseminate to this group. They are receptive and, providing that they know that the report exists, you do not have to put much effort into selling it to them. Sad to say, for most projects, this first group is unfortunately never very large.

The second group consists of the people and organizations that will find something of value and interest in the report but who will not go out of their way to find a copy. You have to sell the report quite hard to this group. You must encourage them to consider the results and, if necessary, to act on the recommendations. Once their interest is aroused, members of this group will want the full report.

The third group may be interested in the subject of the research and might want to know a little about your work and what you found out. They will not really want to bother with the full report but could well be sufficiently interested to read a journal article based on it or to listen to a conference presentation. It is possible that, once they learn a little more about the research, they will follow it up and could well end up looking at the full report. But you will have to work quite hard to get them involved.

Bell, Judith (2005) Doing Your Research Project – a guide for the first time researcher in education, health and the social sciences, 4th edn, Open University Press.

Aimed at students, the book covers much the same ground as this one does.

I shall leave you to judge which does it better.

Bernard, H Russell (2000) Social Research Methods: qualitative and quantitative approaches, Sage.

All you ever wanted to know - and more.

Black, Thomas (1999) Doing Quantitative Research in the Social Sciences: an integrated approach to research design, measurement and statistics, Sage.

A good introductory text that aims to cover the whole research process. It is aimed more at students than at practitioners.

Blalock, Hubert M (1979) Social Statistics, 2nd edn, McGraw-Hill. All you ever wanted to know about statistics but were afraid to ask. First published in 1960 and revised in 1979, this is still a key text for people interested in the statistical side of social research.

Bryman, A (2004) Social Research Methods, 2nd edn, Oxford University Press. A very thorough, well thought-of basic guide to research. It is aimed mainly at students but there is much here for practitioners. There is a companion website.

Creswell, J (2003) Research Design: qualitative and quantitative approaches, 2nd edn, Sage, Aimed mainly at students.

Cutts, Martin (2004) The Oxford Guide to Plain English, Oxford University Press.

An excellent book by Britain's foremost exponent of plain writing. There is much practical guidance here for people producing questionnaires and for those who want to write reports that people can understand.

De Vaus, D A (2001) Surveys in Social Research, 5th edn, Routledge. This is rather academic in tone, but it provides a very full and authoritative guide to the subject.

Fink, Arlene and Kosecoff, Jacqueline (1998) How to Conduct Surveys: a stepby- step guide, 2nd edn, Sage.

A very practical guide to survey design and conduct.

Foster, Jeremy (2000) Data Analysis for Using SPSS for Windows Version 8.0–10.0: a beginner's guide, Sage.

A down-to-earth guide to the most popular statistical analysis package.

Gilbert, Nigel (ed) (2001) Researching Social Life, Sage.

An excellent collection of introductory papers out of the stable at Surrey University that produces Social Research Update. It is both authoritative and practical.

Gorman, G E and Clayton, Peter (2004) Qualitative Research for the Information Professional: a practical handbook, 2nd edn, Facet Publishing.

Not all research is about collecting new data – or primary data as it is known in the trade. A great deal can be achieved by working with data that have already been collected and processed by others. Indeed, most good research begins with a review of what has gone before. This type of research is often referred to as desk research.

Some projects are solely concerned with desk research, relying entirely on the re-analysis of other people's research or on secondary analysis of data that have been collected by others. Even the research that is based on the collection of primary data usually has an element of desk research built in. Few researchers, for example, feel able to manage without some form of literature review or contextual work to position their research.

Desk research covers a range of activities. Literature reviews are the most common. Increasingly the term ‘literature’ needs to be expanded to include material found on the internet. Closely allied to these reviews, and growing in importance, are research reviews which focus on the analysis of actual research findings from a number of different studies. There is also secondary analysis of data where the focus is firmly on the reworking of existing data sets to develop new insights into issues.

This chapter concentrates on the collection of the material used in desk research. The analytical techniques will be dealt with in Chapter 11.

Literature and internet searching

This is a very important part of nearly all research projects, yet it is something that is often dealt with superficially.

No research project exists in isolation. Each piece of work relates in some way to the environment within which the research takes place, to the theories and concepts that have been developed to explain the environmental conditions and to other research on the topic. If your work is to have coherence and relevance you should take full account of what has gone before and what is going on around you. You therefore need to make sure you are fully aware of all the relevant literature on the subject.

The big moment has arrived. After all your careful planning and preparation, you have received approval for the work or the funder has agreed to make the money available. Now you have to get the project on the road and, somehow, keep it going in the right direction.

The project plan

The first thing you should do is to turn the information that is contained in the proposal into a project plan – a document that will set out what you are going to have to do between now and the end of the research.

If you have a well worked-out proposal, this stage is not difficult. If you have been set a firm deadline by which time the project must be completed, then work backwards from this. You know how much time you planned to spend on different tasks, so you can now actually set dates for their start and completion. Now that you are working with actual dates, as distinct from notional weeks or months, you can begin to make allowances for things like public holidays – if your project runs over the winter, you will probably need, for example, to allow at least two weeks for the Christmas period.

If you do not have a fixed deadline, you can begin from now and work forward. You may not be able to start immediately. You may have to recruit staff, or to wait for them to finish work on other projects. It is, however, well worth fixing an actual start date as soon as you can and then to work towards it. Then pace things out from there. Remember to make allowance for things like the time needed to complete and return questionnaires. Check also that you are not committing yourself and the project staff to more days in a month than can be achieved.

Work out the critical path for the project, if you have not already done so for the proposal. This involves looking at each of the tasks and seeing whether they can be carried out concurrently or whether one needs to follow on from the other. This can be important if you are using sub-contractors – they will need to know when they can get started.

The overall shape of the project will be determined to a large extent by the aim and objectives. In the example of research into training and organizational change, a decision had to be made early on about whether to use a before-andafter approach or one that relied on comparisons between organizations. And so it is with most projects. Even though the overall shape may be fixed by factors outside your control, there is still much to be done to design and plan the study in detail.

Resource constraints

First it is necessary to be aware of the constraints on the resources available to you. Essentially you need to think about money, time and expertise. Each is equally important.

Money

The overall amount of money available will play a large part in determining the design of the research. If you are being contracted by an organization to carry out work on their behalf, you clearly have to work within the budget they set. Research is a labour-intensive process and so most of this budget will be taken up by staff costs. Basically, the amount of money available dictates the number of days that can be spent on the project and thus determines its overall size.

If the research is part of your job within an organization there is usually a less stringent financial constraint. Most of the staff costs are likely to be covered, unless it is necessary to employ additional researchers, and the budget will be used for incidental expenses like the costs of conducting a survey. However, financial resources are never unlimited and before settling on the design you should check that money is available to spend on the research.

At the other end of the scale, if you are a student, it will probably be a case of considering whether or not you can afford the postage, copying and stationery costs associated with undertaking a postal survey.

Time

Here you need to think about two different sorts of time: elapsed time and research days. Elapsed time is simply the time between the start and finish of a project. For a student, the deadline is the submission date for the dissertation. On a research contract there will almost certainly be an equally severe cut-off point. In-house researchers may have their deadline set by something like the date of a board meeting.

No research project is complete without some form of report. Without a report, the only people to benefit from the work are the researchers.

The nature of the report should be determined by the project itself and by the audience to which the report is addressed. Short projects carried out within an organization may simply be reported by means of a brief paper supported by an oral presentation. Academic research, on the other hand, is expected to result in lengthy dissertations which cover all aspects of the research and report on them in a rather formal, predetermined way. Between these two extremes there is considerable scope for variety.

Structure

Most reports, no matter what their size or formality, follow a common basic structure. You must say why the work was done, what events led up to it and what other work was relevant. You must state the aim and objectives and describe what you did in order to achieve the objectives. This leads you through to the results, the conclusions you draw from the results and then to the recommendations.

These elements represent a common basic framework that you can start with when planning the structure of your report. You will almost certainly have to deviate in some way from the basic structure but it will provide you with a starting point.

Building up an outline

The trick about writing good reports is to think things through in advance and to build up a structured outline before beginning the actual process of writing. Before you start writing the first paragraph you should know how many chapters the report will contain, what will be the main and subheadings within each chapter and what material will be put in appendices. You should know roughly how long the report will be and what it will look like at the end of the day. You should also have a very clear idea of the audience for which you are writing.

You should ask yourself what sort of people they are, what prior knowledge of the subject they will bring to the report, how long they are likely to spend looking at it and who they might pass the report on to.