Negative correlation Essay Example for Free

Negative coefficient of correlation EssayHowever, the weakness of this cast out correlation raises the notion that the results have no correlation at all. To investigate this possibility I will first try to find how just how weak this negative correlation is. As you bottom see from graph 2 the different result values for each temperature vary quite a lot from the average. This undermines the verity of the results, as the greater the range of the results from the average the less carely the results and therefore the average is accurate. The range of the different results for each temperatures and the standard deflexion from the mean notify indicate how accurate the results are. This is be coiffure, as I have explained to begin with the range and the deviation of the results from the mean determines how accurate the results are. Graphs 3 and 4 fate how the standard deviation and the range of the results change over the temperature range used in the experiment.In graph 3 th e range was calculated by subtracting the lowest value from the steepest, and in graph 4 the standard deviation was calculated using the formula ? ((? (x -? x)/n) (where x = the values and n = the number of values). Graph 3 and 4 show that in general the accuracy of the result increases with an increase in the temperature, and that at low temperatures the results are not very accurate at all.I can conclude this as, the bigger the standard deviation or range the less accurate the results are. In both graphs 3 and 4 the general pattern shows high range and high standard deviation at low temperature (high inaccuracy) and low range and low standard deviation at high temperatures (low inaccuracy). This therefore contradicts my earlier theory that the inaccuracy would increase with an increase in the temperature, barely it must be taken into account that the results are very inaccurate already.Because of this inaccuracy the results may have had a certain random element in them, evidence for this is in the fact that the results are very inaccurate (they have a large range and a large standard deviation), and that the inaccuracy varies a lot from its trend line (see graphs 3 and 4). So in conclusion, I cogitate that the results of this experiment were not accurate at all and not accurate enough to draw both sound conclusions that are backed with scientific theory/ knowledge.In fact I think there was a certain random element in the results that contributed to the inaccuracy of the results. Because of the inaccuracy of the results, I feel that it is not a good idea to use these results as a basis for any conclusions, especially as these conclusions cannot be backed up by any scientific knowledge. Therefore with only the results I have at this moment, I have to conclude that the resistance is directly proportional to the temperature. I can conclude this because I have decided to use scientific knowledge as a basis for the answer to this correlation.I have done this b ecause the results I have are too inaccurate to use for making any judgements on this correlation. Evaluation So far I do not think this investigation has gone very well because of the very inaccurate results, however with the large summation of scientific knowledge I have been able to come to a conclusion to this investigation that has a lot of backing. My prediction was of corse correct. The land to this inaccuracy in the results is, in my opinion due to the large amount of contacts and wiring used in this method.Some components react to temperature by changing the resistance in different ways, most components react in a directly proportional linear way, but some components like a filament lamp or rectifying tube react in a directly proportional but logarithmic way and/or only at certain voltages. It is quite likely that in the multi-meters there were some of these components that react in a logarithmic way and at certain voltages. This means that the current flowing through th e circuit would be affected out of proportion by these components, therefore affecting the resistance.Although the affect of these components would have been neutralised if one multi-meter were used, it might not be neutralised if both multi-meters were used as they were in this method. The large amount of wire used in this experiment has meant the current could have been easily affected by separate variables that I did not take into account, like a change in the room temperature, or if I accidentally spilt some water onto the wires. Also like the different depths used in the water bath, or a magnetic flux from the high voltage used in the kettle.Due to the large amount of exterior variables that can affect the large amounts of wire used, I have to conclude that the large amounts of wire used in this method was the main cause of the inaccuracy of its results. I cannot estimate how accurate my results are because I have no basis on what the true values are, only that the correlati on is directly proportional. If I were to repeat this investigation, I would change the method so that less wire was used and only one multi-meter was used, I would also change the method so as to stop any of the variables mentioned in this evaluation affecting the results.If possible I would try and use a more accurate multi-meter in my experiment. A more accurate multi-meter will give me more accurate results the reduced amount of wire to be used will reduce the effect of outside variables and make the results more reliable. I dont think I need to do any extra research, as I have more than enough scientific backing for any conclusions I will have to make if I repeat this investigation.