How should I structure the Conclusions?

The Conclusions section is not just a summary. Don’t merely repeat what you said in the Abstract and Introduction. It is generally not more than one or two paragraphs long. A Conclusions section typically incorporates one or more of the following:

1. a very brief revisit of the most important findings pointing out how these advance your field from the present state of knowledge

2. a final judgment on the importance and significance those findings in term of their implications and impact, along with possible applications to other areas

3. an indication of the limitations of your study (though the Discussion may be a more appropriate place to do this)

4. suggestions for improvements (perhaps in relation to the limitations)

5. recommendations for future work (either for the author, and/or the community)

6. recommendations for policy changes

 

HOW to begin the Conclusion:

Just as professional copy editors advise against beginning a paper with This paper

describes, they also suggest avoiding ending the paper in the same way (This

paper has described). This is for three reasons:

• they waste a lot of words (5–7 words that tell the reader nothing)

• they delay the main topic

• they are not memorable for the reader and have no impact

 

Possible ways to start the Conclusion:

- Compression plays an important part in … In fact, it was found that …

- A number of compounds present in X are responsible for delaying the onset of …

- The crystal structure of X reveals that …

- The localization of X in neurons suggests that it is a good marker for neuronal viability.

 

If you don’t have any clear conclusions, it is important not to present your findings in an exaggerated light or to say something uninteresting or irrelevant.

Here are some examples of authors admitting that their work did not achieve all that they had hoped for. In some cases readers are immediately warned of this ‘failure’ through the use of the words highlighted in italics.

 

Unfortunately, we could not assess how much of the difference in outcome was due to..

When results are compared across different components, the confidence intervals overlap, and we have no conclusive evidence of differences in...

Although some progress has been made using our model, this incremental approach provides only a partial answer

Unfortunately this trial had too few subjects to achieve sufficient power and had a low …

It is also unclear what conclusion should be drawn …

Regrettably, we did not have the means to …

To make your Conclusions not sound too negative, you can add some hope for the future.

Although it is too early to draw statistically significant conclusions, two patterns seem to be emerging …

However, more definite conclusions will be possible when...

Nevertheless, our study confirms recent anecdotal reports of …

Despite this, our work provides support for …

In any case, we believe that these preliminary results indicate that …

 

HOW to end the Conclusion

 

However, it remains to be further clarified whether our findings could be applied to …

Further studies are needed to determine whether these findings could be applied to components other than those used for …

One area of future work will be to represent these relationships explicitly …

Future work will mainly cover the development of additional features for the software, such as …

Future work will involve the application of the proposed algorithm to data from …

Future work should give priority to (1) the formation of X; (2) the interaction of Y; and (3) the processes connected with Z.

Future work should benefit greatly by using data on …

We suggest that policy makers should give stakeholders a greater role in …

We recommend that stakeholders should be given a great role in …