Kmahal1990
Member
- Joined
- Mar 20, 2006
- Messages
- 49
- Gender
- Male
- HSC
- 2007
Reliability is determined upon:
1. How many times the experiment has been repeated, and how precise (near each other) are the results. The greater the amount of times you repeat the experiment, and the more precise the results, the more reliable the experiment is.
To work out precision, find the average of your final results* and find the result that differs greatest from your mean. The result is your GDM (Greatest deviation from the mean).Divide the difference of your GDM and the mean of all your results, and then times that by 100, to give you the precision of your results. It is a percentage value. For an experiment to be reliable in this respect, the degree of precision should be smaller than 10-15%(and maybe lower for public schools, lol).
*By final results, i mean not the separate measurements you take during the experiment, but the result of what you intend to find, in each case of the repeated experiment. For example in the pendulum experiment, your final result for each repeated experiment would be the value for gravity according to the measurements you have taken, not the single measurements of time taken for oscillations.
2. How well you've kept your variables constant in your repeated trials
Example: If you are working out the value of gravity from the pendulum experiment, swapping the stopwatch you use, and or swapping the person who times the experiment will lower the reliability and the validity in this case. A different timer has a different response time, and therefore will lower the precision of your values.
3. How precise the measuring devices you use are. For example, results obtained using a ruler separated into centremetres and milimetres is more reliable than a ruler with only centremetres.
4. Your result is more reliable, if you use a graph and a line of best fit to obtain the result. A line of best fit placed on a graph along with your values identifies outrider values( values that are completely different to every other value) and rule out systematic error in some cases.
Validity is determined by:
1. First and foremost, accuracy. The accuracy of your experiment, is how different your result is from the result you are meant to get theoretically (the ideal result). To find accuracy, find the difference between your result and the ideal result, then divide it by the ideal result, times by hundred (expressed as a percentage). For example, your result for gravity from the pendulum experiment is 8, and the ideal value for gravity 9.8 (lets say) then the accuracy would be:
(9.8-8)/9.8 X 100 = 18.4%
For an experiment to be considered valid in this respect, the degree of accuracy should be less that 15%.
2. Keeping variables constant also affects validity.
3. Your method must be valid, in that it tests your hypothesis and only your hypothesis, and is not affected by effects that cannot be measured and taken into account.
4. Your results are reliable.
In your conclusion, you should also mention what reasons may have caused your result to be different from the ideal value. Sometimes these are errors, and sometimes they are not your fault ( for example, the value for gravity is different on different parts of earth. So obviously your gravity will differ slightly to the gravity in books.
Errors are split into two categories, systematic and random errors. Systematic errors are errors that occur on every repetition of the experiment. For example if your electronic scale is off by .5 of a gram, then the value of all your experiments are off by .5.
Random errors are errors caused by a variation in method of measurement. (For example on 2 occasions you read the volume of a breaker properly, and on the third you read it wrongly because of paralax error).
In your conclusion, you must also mention what ways you can minimise your errors and improve your method.
Remember to get straight to the point, it is about presenting the findings of your experiment. This experiment is reliable/unreliable because the degree of precision of the results is 3%/50%. This measure here was taken, and so was this to make the results more reliable. This experiment is invalid/valid, as the accuracy of the experiment is 3%/50% and because the results were reliable/unreliable. This deviation from the theoretical value is caused by this, this and this. They can be reduced by doing this, this and this. Thats all there is to it.
There is probably lots that i missed out on, feel free to contribute.
1. How many times the experiment has been repeated, and how precise (near each other) are the results. The greater the amount of times you repeat the experiment, and the more precise the results, the more reliable the experiment is.
To work out precision, find the average of your final results* and find the result that differs greatest from your mean. The result is your GDM (Greatest deviation from the mean).Divide the difference of your GDM and the mean of all your results, and then times that by 100, to give you the precision of your results. It is a percentage value. For an experiment to be reliable in this respect, the degree of precision should be smaller than 10-15%(and maybe lower for public schools, lol).
*By final results, i mean not the separate measurements you take during the experiment, but the result of what you intend to find, in each case of the repeated experiment. For example in the pendulum experiment, your final result for each repeated experiment would be the value for gravity according to the measurements you have taken, not the single measurements of time taken for oscillations.
2. How well you've kept your variables constant in your repeated trials
Example: If you are working out the value of gravity from the pendulum experiment, swapping the stopwatch you use, and or swapping the person who times the experiment will lower the reliability and the validity in this case. A different timer has a different response time, and therefore will lower the precision of your values.
3. How precise the measuring devices you use are. For example, results obtained using a ruler separated into centremetres and milimetres is more reliable than a ruler with only centremetres.
4. Your result is more reliable, if you use a graph and a line of best fit to obtain the result. A line of best fit placed on a graph along with your values identifies outrider values( values that are completely different to every other value) and rule out systematic error in some cases.
Validity is determined by:
1. First and foremost, accuracy. The accuracy of your experiment, is how different your result is from the result you are meant to get theoretically (the ideal result). To find accuracy, find the difference between your result and the ideal result, then divide it by the ideal result, times by hundred (expressed as a percentage). For example, your result for gravity from the pendulum experiment is 8, and the ideal value for gravity 9.8 (lets say) then the accuracy would be:
(9.8-8)/9.8 X 100 = 18.4%
For an experiment to be considered valid in this respect, the degree of accuracy should be less that 15%.
2. Keeping variables constant also affects validity.
3. Your method must be valid, in that it tests your hypothesis and only your hypothesis, and is not affected by effects that cannot be measured and taken into account.
4. Your results are reliable.
In your conclusion, you should also mention what reasons may have caused your result to be different from the ideal value. Sometimes these are errors, and sometimes they are not your fault ( for example, the value for gravity is different on different parts of earth. So obviously your gravity will differ slightly to the gravity in books.
Errors are split into two categories, systematic and random errors. Systematic errors are errors that occur on every repetition of the experiment. For example if your electronic scale is off by .5 of a gram, then the value of all your experiments are off by .5.
Random errors are errors caused by a variation in method of measurement. (For example on 2 occasions you read the volume of a breaker properly, and on the third you read it wrongly because of paralax error).
In your conclusion, you must also mention what ways you can minimise your errors and improve your method.
Remember to get straight to the point, it is about presenting the findings of your experiment. This experiment is reliable/unreliable because the degree of precision of the results is 3%/50%. This measure here was taken, and so was this to make the results more reliable. This experiment is invalid/valid, as the accuracy of the experiment is 3%/50% and because the results were reliable/unreliable. This deviation from the theoretical value is caused by this, this and this. They can be reduced by doing this, this and this. Thats all there is to it.
There is probably lots that i missed out on, feel free to contribute.
Last edited: