Reporting and critical evaluation of biological research

As you read through this final section, remember that this is all highly relevant to your final AH Biology Project Report. You can find examples of previous candidate AH Biology Projects on the SQA Understanding Standards website and you can use the Marker Commentary to reflect on each candidate's performance. Use the information in this section to fuel your own understanding of what should be written in a final Project Report.

• An abstract. This includes your aim(s) and a summary of your main findings. The aim must link your independent and dependent variables clearly. Note: In the example of a Title above - "caffeine" is NOT the independent variable - it would likely be something like "caffeine concentration" - you must be much clearer in your aim.

Below, is an example of an abstract I have taken from one of the published projects on the SQA Understanding Standards website. The full project can be found here. It is impossible to assess this without reading the full project (note this should be one of the final things that any researcher writes since it must summarise overall findings) - however, the abstract is found immediately after the contents page and includes an aim and accurately summarises the findings of the investigation.

A note about Citing and Referencing sources

Any body of writing produced by a researcher that reflects wider reading and information gained from others must be referenced fully. Referencing a source involves:

1. Providing an in-text citation

2. Providing a full reference list at the end of the document.

There are 2 Referencing systems that SQA have requested you use for your own projects - either the Harvard referencing system or the Vancouver referencing system. I have included some examples from the world of research below but have gone into much more detail about this in the following YouTube video.

(b) Reporting and evaluating experimental design

The Procedures section of an experimental report should reflect a lot of considered planning. This is a fundamental foundation to your Project (or any piece of research). I have written this section as numbered bullet points that reflect the SQA's marking instructions for the procedures section of your final Project (although you should keep in mind that this is how any reporting and evaluation of experimental design should be performed).

A Procedures section, reporting and evaluating experimental design should have the following features:

1. An overall methodology that allows the aim and hypothesis to be tested, otherwise the design is invalid.

2. A method section that contains sufficient information to allow another investigator to repeat the work. This should be written in past tense and impersonal tone, as you will have practised throughout high school. An important feature here is that EVERYTHING is described, e.g. how were solutions prepared (a technician should not be preparing AH Biology level solutions unless a Risk Assessment demands this support), what apparatus were used to make measurements, etc.

3. A rigorous evaluation of the validity and reliability of the experimental design. This involves identifying suitable controls (negative and positive where appropriate - as discussed in a previous section, this is not always appropriate so please refer back if you have forgotten). This also relies on a thorough examination of any confounding variables. The validity of the experiment may be compromised when any factor other than the independent variable could influence the value of the dependent variable. Consider how these confounding variables will be controlled or, if this is not possible, how will they be monitored throughout the experiment. Reflect back on randomised block design if particular confounding variables cannot be controlled.

4. A clear demonstration of representating sampling - this includes the effect of selection bias and sample size. Selection bias is the selection of a sample in a non-random way, so that the sample is not representative of the whole population. Sample size may not be sufficient to decide, without bias, whether the change to the independent variable has caused an effect in the dependent variable.

5. An independent replicate - as discussed previously, results can only be deemed reliable if they can be independently replicated. In this section of a report, a candidate must describe how the independent replicate was carried out.

6. Justification of how the pilot study informed the final methodology - remember, prior to setting up an investigation, there will be lots of initial questions. Your teacher will not provide you with a protocol for your investigation. You might not know how much substrate to add in an enzyme reaction, or what concentration of caffeine to incubate a Daphnia safely in, or for how long to record monkey behaviour in the zoo. These initial questions are investigated in a Pilot Study and used to inform your final procedure.

7. Finally, your Procedure section should include a methodology that reflects the Advanced Higher level you are working at. Go big or go home! Do your research - you might have done a bog-standard basic protocol in S4 and want to relive it now - but don't! Read around and see if you can adapt your methodology to improve accuracy of results.

The summary of SQA Procedure Checklist is below:

(c) Data analysis

Having carried out a procedure, a set of results will be generated. These must now be analysed, perhaps involving the calculation and presentation of the mean, median or mode. Let have a quick look at the difference between these terms.

According to Ruxton and Stafford (2015) in their SSERC Statistical Guide for AH biology projects, "a good rule of thumb would be to use the mean to describe the typical value". If the data seems asymmetrical (many values lying far from the mean at the upper or lower end), then the median may be more appropiate. A quick way to check if your data is approximately symmetrical (thus making the mean the best option to describe your data), is to calculate the mean and median values: if they are similar, your data is approximately symmetrical.

You can use R to do this. Consider the following data below, looking at catalase activity in the presence or absence of inhibitors - this is measured by the volume of oxygen produced:

The coding used in R to create summaries of this data set is shown below:

Standard deviation and Range

While the mean and median are common ways to describe the central tendency of the data, the spread of the data can be described using the range or the interquartile range (as discussed in KA2). The interquartile range describes the spread around the median value.

However, the standard deviation describes the spread around the mean in a roughly symmetrical data set. The lower the standard deviation value, the smaller the spread of data from the mean.

Assessing Statistical Differences between Samples

Statistical tests are used to determine whether the differences between the means are likely or unlikely to have occurred by chance. A statistically significant result is one that is unlikely to be due to chance alone.

A statistical test for a difference between two samples

A t-test can be used to compare the differences between the means of two samples relative to the spread of their values, assuming the distribution approximates to "normal" (as discussed above). If the spread of their values between the two samples overlaps, then the difference between the two means is considered less likely to be significant (and more likely due to chance).

Lets look back at the data comparing the effect of aspirin and lead nitrate on catalase activity. The null hypothesis for this investigation would be that the tested inhibitors would have no effect on catalase activity - the t-test compares the means and the p-value generated relates to this null hypothesis.

If the p-value is >0.05, we accept the null hypothesis and say that there is >5% chance that the difference between the means occured by chance alone.

If the p-value is <0.05, we reject the null hypothesis and can conclude that we have statistically significant evidence that the inhibitor affects catalase activity.

However, the 5th line of code instructs R to perform a t-test between the negative control ("noinhib") and the aspirin data. A p-value of 0.09 has been reported. Since this is >0.05, we must accept the null hypothesis that aspirin has no statistically significant effect on catalase activity.

A wealth of statistical support is available via the following SSERC booklet, written by Graeme Ruxton and Jim Stafford. It is downloadable via this link.

(d) Evaluating results and conclusions

Following the presentation of results, it is time to bring everything together and form a Conclusion. The conclusion reported must refer to the aim and hypothesis and be supported by the results documented throughout the report. Be cautious here to refer to the genuine dependent variable, e.g. when measuring catalase activity, it is probable that you may be actually measuring volume of oxygen produced rather than "catalase activity" - Your conclusion must refer to catalase activity and not to volume of oxygen produced.

The validity of any conclusion drawn relies heavily on how rigorously the experimental section of a project was planned. A valid conclusion can only be drawn if:

• Confounding variables have been satisfactorily controlled.

• A suitable sample size was used for the investigation.

• The whole investigation was independently replicated.

Validity and Reliability of Experimental Design

The validity and reliability of the experimental design should be taken into account. Consideration should be given as to whether the results can be attributed to correlation to causation. Have a look at Task 19 for some more practise.

There are two useful outputs to note:

1. The p-value - this value is associated with the null hypothesis, which in this case is that there is no association between age and reaction time. Since the p-value is <0.05, we can reject the null hypothesis and conclude that the two variables are related. But what is the strength of this relationship?

2. The last value on the output report is important here - this is the Pearson's product moment correlation coefficient (or Pearson's r for short). Pearson's r can range from -1 to 1. A positive value (0-1) indicates a positive correlation while a negative value indicates a negative correlation. The closer the value is to 1 or -1, the closer the points on the scatterplot would be to a straight line of best fit - i.e. the stronger the correlation between the two variables. Our value of 0.93 suggests a strong, positive correlation between age and reaction time.

Evaluation of Results

Evaluation of conclusions should also refer to existing knowledge and the results of other investigations. Meaningful scientific discussion would include consideration of findings in the context of existing knowledge and the results of other investigations. Scientific writing should reveal an awareness of the contribution of scientific research to increasing scientific knowledge, and to the social, economic and industrial life of the community.