Prediction summary
I will expect to gain different results for each sample. There are a number of factors which will lead to this. From the information of the samples above I can predict that,
- Sample 1 will travel the distance the slowest
- Sample 2 will travel faster than sample 1 but not quite as fast as the blood from sample 3.
- Sample 3 will travel the fastest.
Pilot data
I used the following method when I did my pilot experiment,
- I filled three measuring cylinders with the 0.1-3 copper II sulphate solution. To above approximately 5cm the 100cm level.
- Using a Pasteur pipette I introduced some simulated blood into the pipette.
- Using fingers to steady the steam I put the tip of the pipette just above the copper sulphate solution.
- By pressing gently on the bulb of the pipette I will release a small drop of the simulated blood into the copper sulphate solution
-
I will allow the blood droplet to reach its average speed before starting to measure the time at the 100cm mark.
- Once the blood reaches the 10cm mark I will record the time taken to travel the distance between the two marks.
- I plan to repeat this method at least 10 times for each blood sample.
These are the results that I obtained,
To work out the standard error of these results I use the formula 5
Se = mean +/- (2.26 x standard deviation / (square root of 10))
Se = 10.77 +/- (2.26 x 2.36 / (square root of 10))
Se = 10.77 +/- 1.69
From the calculation on page 4 I am 95% certain that the mean of the readings in the results lies within the range 10.77 +/- 1.69 which means that my results are reasonably accurate and precise. The standard error of my results proves that my method works and can achieve fairly accurate results. However, ideally the standard error would be below 1 if I had really consistent results. This was just a pilot experiment with the aim of practising my method. The graph I have drawn shows that I had 2 extreme values (values circled). These could be extreme values for many reasons.
- The blood could have dispersed
- The blood droplet may have had a larger mass than the others
- The blood was released too low.
When I perform the real experiment I will have a better idea of what I am doing and the techniques needed to gain more precise results. The method I used was fine however I feel one modification needs to be made.
In my method I stated that I put the tip of the pipette just above the copper sulphate solution. However, I feel that it would be better to position the pipette just under the solution. This may help to stop the blood sample hitting the sides of the test tube and separating.
Apparatus required for practical
Variables that could affect the results of the experiment
- Size of blood droplet – When using a Pasteur pipette it is very hard to produce droplets of simulated blood of the same volume. For this reason volumes will vary everytime. However, this can be partially controlled by applying the same pressure to the ‘bulb’ on the pipette until a droplet forms of a similar size to the other droplets.
- The placement of the Pasteur pipette must be the same each time, though this is very hard. I will place the tip of the pipette just below the surface of the copper sulphate solution. By doing this each time I release a droplet it will help to keep the variable controlled.
-
When I release a droplet of blood I must begin the timing at the 100cm level and stop the timing at the 10cm level. To make sure I do this accurately I must minimise parallax error. To do this I must adjust myself so I am at eye level when I start and stop the clock at the markings. This means that it will be more likely that I will stop and start the clock at the correct time for each blood sample.
-
The acceleration of the blood droplet – The blood droplet must be given sufficient time to accelerate to its terminal velocity under gravity before I start the timing. To counter this I will add an additional 5cm3 of copper sulphate solution to the beaker. Therefore this will allow the droplet to accelerate before it passes the 100cm3 mark and I start the stop clock. This method should let me partially control the variable and will let me produce a set of fairer results.
- Air bubbles in the Pasteur pipette – If there are air bubbles in the Pasteur pipette when the simulated blood is taken up the blood droplet may disperse when it is released and it will give an inaccurate measure of the time taken to travel the total distance. To eliminate this problem I will have to make sure there are no air bubbles taken up with the samples.
To improve my test I could increase the accuracy of my results. This could be done by improving my equipment. To improve the fairness of the test I could release a measured amount of blood into the solution so that all blood droplet amounts are constant, this would produce fairer results. With additional equipment I could also make sure that the pipette was positioned at exactly the same place each time I release a blood droplet, again aiding fairness of results.
For each blood sample I will be performing 10 separate observations and results. Increasing the amount of observations generally means that my experiment will be more accurate and produce more reliable results. My results should be fairly constant and will enable me to easily judge whether there are any extreme values in my sets of data. I should expect some extreme values but I will be able to tell generally how accurate my methods have been if there is a lot of variation in my results. Obtaining 10 results per sample should be sufficient enough to represent a larger data set providing there is not a great amount of variation.
I will use standard deviation and standard error to asses the reliability and precision of my results. Confidence limits give a measure of the accuracy of the mean. They help me to understand that the true mean of a set of data lies within a certain distance of the mean. I am able to say that I am 95% certain that the true mean of my results in the data lies within a certain range.
Safety
6
Results
My raw data is located in the appendix of my coursework, my processed results are shown below.
Relative densities
To compare densities I will assume the relative density of the normal sample is 1.00
The relative density of unknown sample is found by formula,
RDsample1 x time sample 1 / time unknown sample.
RD sample 1 = 1.00 x 10.25 / 8.03 = 1.28 (28% increase in density over sample 1)
RD sample 2 = 1.00 x 10.25 / 6.49 = 1.58 (58% increase in density over sample 1)
Tally chart
A tally chart helps to decide if the data is normally distributed by plotting a frequency histogram using the chart. The intervals help to show how compact the data is, they are found by dividing the difference between the maximum and minimum values by ten. I have rounded the intervals to the nearest tenth.
Key: | represents a count of one / represents a count of five.
Maximum value 14.02
Minimum value 8.66
Difference = 5.36
Intervals = 5.36 / 10 = 0.536 = 0.5cm intervals
Maximum value 10.42
Minimum value 66.63
Difference = 3.79
Intervals = 3.79 / 10 = 0.379 = 0.4cm intervals
Maximum value 7.69
Minimum value 5.74
Difference = 1.85
Intervals = 1.85 / 10 = 0.185 = 0.2cm intervals
Through these tally tables I have been able to draw a frequency histogram. The histograms give a visual image of the distribution of data values - where they are clustered and/or how widely spread out they are.
Conclusions
After analysis of my results I am able to draw conclusions about the blood samples, I will question the reliability of my results in my evaluation.
-
Sample 1 – Sample 1 has the highest mean value (10.251/s). This means that this sample contained the smallest haemoglobin mass and was the smallest relative density. This sample can be confirmed that it was taken from the healthy male who had not been blood doping or participated in regular exercise. Because of this the bone marrow did not make more red blood cells and the red blood cell count remained ‘average’. Therefore he contained the normal amount of haemoglobin and this resulted in his blood having the lowest relative density. This is why this sample travelled the standard distance the slowest. My conclusion for sample 1 is valid because from my scatter graph I can tell that the results for sample 1 are fairly compact and most are close to the mean. There are 3 results slightly above the mean, however I have not discarded these because without them I would not have a very large sample.
-
Sample 2 – Sample 2 had a mean value of 8.039/s this value was between that of sample 1, the healthy male, and sample 3 the fastest sample. This suggests that this sample was taken from the healthy male after 6 months of regular exercise. The 6 months of regular exercise would have increased the amount of red blood cells in the male’s blood and this resulted in an increased relative density of blood. Therefore his blood sample fell faster than sample 1 because it contained more haemoglobin but not as much as sample 3 seems to have had. Because of this it is obvious that this male had not used blood doping, but has become accustom to exercise and therefore stimulated an increase in steady red blood cell production. Sample 2 had a 28% increase in relative density from sample 1 which supports my hypothesis. My scatter graph shows that my results for sample 2 are fairly unevenly distributed, they are not as precise as my results for sample 1. Most results though are still around the mean.
-
Sample 3 – The mean value of this sample was 6.487/s, and the relative density was 1.58. Therefore this blood sample must have been taken from the healthy male after suspected blood doping. The blood doping would have increased his red blood cell count more dramatically than regular exercise (sample 2) and this resulted in the largest relative density per blood drop, this is why the blood travelled the standard distance the fastest. The mean value is just under 4 seconds faster than that of sample 1 and this supports the dramatic increase in the density of blood confirming that the male must have been blood doping for such a great increase compared to sample 2. The 58% in relative density increase over the normal sample 1 is far too dramatic to have been triggered by exercise. However, my background scientific knowledge states that successful blood doping can increase the haemoglobin level and RBC count by up to 20%. 7 I have a 58% increase which is far more dramatic than expected. However, this could be due to two factors. Simulated blood may not act like real blood or the data may no be completely valid as supported by the lack of similarity to a normal distribution. The scatter graph for sample 3 shows a fairly even spread of data between the mean value, data is present at the mean and this supports the validity of my results.
My hypothesis has been proved, I expected to see an increase in density of blood in samples 2 and 3 and my results support this as the mean travel time for the samples 2 and 3 are faster than sample 1 for almost the same volume for each drop of blood.
Evaluation
My experiment was fairly successful as I was able to draw valid conclusions from my results. However, there are procedures I could have taken to improve the precision and reliability of my evidence. My evaluation discusses this.
Anomalous results
Below is my table of results – I have highlighted the potential anomalous results.
If there were to be any anomalous results then they would be the ones highlighted. These are probably anomies because they occurred in my first sample at the beginning of the experiment and it took me a while to master the technique. To overcome this I could have done a few test drops before I recorded the actual results. However, they are within two standard deviations of the mean therefore they are not that dramatic and I have not excluded them because they do not affect the outcome of my conclusion or outcome of my results greatly. It is likely that these results were affected by errors in my procedure.
General errors and errors in procedure
In some of my results I encountered errors which were due to a number of reasons. Before my experiment I setup a table of errors so I could record any errors I encountered in my experiment and recall them in my evaluation on the basis that they have affected the outcome of my results slightly.
Accuracy and reliability of results
From my results I drew frequency histograms for each blood sample. By using a histogram you can see how wide distribution of data values are.
For blood sample 1 the histogram shows that the data is relatively clustered, most results are around the 8.5 – 9.5 /s group this histogram most closely resembles a normal distribution which shows that the data is precise and valid.
Sample 2 histogram has no clustered data and does not resemble a normal distribution with little data around the calculated mean – this is less precise data than sample 1.
The histogram for sample 3 is widely spread out, there are no real clusters of data, however there is data present at the calculated mean.
To have improved precision it would have been better to take more results. If my investigation was error-free my results would have been more likely to be consistently closer to the mean value. I would have recorded a minimum set of 20 results had time allowed.
Main source of error
The main error I had was the size of the blood droplet released. It was extremely hard to maintain a constant volume of blood. When making droplets of blood I had to press the bulb on the pipette very lightly to allow a droplet to form at the tip of the pipette – this was tricky to do. The error also resulted in other consequences that affected the accuracy of my results; if I pressed a little too hard on the bulb then some blood would come out and just drop into the solution which would cause it to disperse. I also found it difficult to keep my hand stable, and any sudden movement would mean the blood droplet would leave the pipette at the incorrect volume.
The error was the main cause of my inaccurate results and imprecision as it affected the time it would take for the blood droplet to travel the distance. It did not affect my conclusion though because the mean values allowed me to establish which blood samples had a relative high density.
Improvements
In the future if I were to perform the experiment again I would make several changes. These improvements to my method would enable me to collect more reliable and precise results.
- With additional apparatus I believe I could have improved the accuracy of my results. If I were to use a clamp then it would have been possible to position the pipette in exactly the same position just underneath the solution each time. This would be likely to eliminate some of the errors I experienced during my investigation like the dispersal of the blood and the blood hitting the side of the measuring cylinder, both increasing the time taken for the blood to travel the entire distance. Using a clamp would have meant my results were more accurate because the blood droplets would have been released from the same position.
- Though this improvement is not an essential one if the above is put into practice I would use a different type of measuring cylinder to drop the blood into. The measuring cylinder would have to be wider, having a wider cylinder would mean that it is less likely that blood droplets would hit the sides of the beaker and slow them down. This would eliminate a factor which increases the time taken to travel the distance and would make my results more accurate.
- Using a Pasteur pipette means it is very tricky to produce a constant volume of blood droplet. By using a different piece of apparatus it would have been easier to make droplets the same volume. Having the same size droplets eliminates this variable and means my results would be more accurate. An ideal tool for getting a constant volume of the blood would be an automated pipette.
- It would have been possible to reduce the parallax error if I had been able to use somebody to stay seated and time the blood droplet to travel the total distance. This person would be dedicated to timing the blood droplet and would not have to worry about releasing the blood droplet. They would just been doing the timing and therefore it is likely that they would have timed more accurately because they would become accustom to where to position themselves to reduce parallax error and because all they would have to do is time the droplet their reaction times would be minimal.
- Using the same pipette for every solution means that when you wash the pipette out water stays in the pipette and this dilutes the blood sample when it is taken up. Diluting the blood sample affects the relative density of the blood sample and therefore to avoid this problem in the future I would have used a new pipette for every drop of blood.
Validity of conclusion
The true means of my samples show some overlap.
This would suggest that the data is not as valid as it could have been. I think that performing at least 20 measurements for each sample would help to validate my conclusion.
Spell Check proof
References
1
2
3
5
6
7
2 REFERENCE: Figures are taken from my lecture notes on red blood cells
3 REFERENCE: Information from an article on blood doping (http://tc.cc.tx.us/~mstorey/beckham.htm)
1 REFERENCE: From lecture notes on red blood cells
4 REFERENCE: information from a PowerPoint about measuring blood, also from the human biology textbook for AS. Pictures fused are from the PowerPoint.
5 REFERENCE: Sheet on how to manipulate raw data and make summaries – this was handed out during a lecture.
7 REFERENCE: Information from an article on blood doping (http://tc.cc.tx.us/~mstorey/beckham.htm)
6 REFERENCE:
A safety sheet about Copper sulphate from Cleapss cd rom