Global Journal of Science Frontier Research, A: Physics and Space Science, Volume 22 Issue 1

c) Data cleaning Outliers are those observations that appear to deviate markedly from the others in a sample, and consequently may be indicators of erroneous data. For example, they may be caused by errors in the experimentation process. If these do represent incorrect information, they should be eliminated, or if possible, corrected. On certain occasions, it may not be possible to identify whether the outliers are the result of errors since they could be caused by random variations of the analyzed event. Various formal tests have been established for the correct identification of possible anomalous values, among which it is assumed that the distribution is normal. However, in the present work, such methods cannot be used because they require a large sample that is considered representative of the complete data set. Therefore, another method was carried out and it consisted of graphing the recorded data and adding its trend line (Figure 5), in order to discard those values that were considerably away from it. In total, four measurements out of the twenty-two captured were discarded. The discrepancy of these was probably caused by various sources of error, mainly by measurement errors associated with the infrared thermometer, such as the inclusion of other heat sources within its field of view and the reflectivity of other bodies. It could also be caused by the sudden or rapid reduction of the voltage, as explained in Table 2, and that in some attempts the electromagnet was heated more evenly, among other factors. VI. C onclusion and E valuation Through the experiment, various sources of error were found that could alter the experimental results, which are explained in detail in tables 7 and 8. Table 7: Identification of possible random errors with their respective correction methods Random error Correction method Instrumental uncertainties. - Repetition of measurements under controlled conditions. - Comparison of the voltage readings produced by the power multimeters. Variation in the heating uniformity of the electromagnet. Selection of experimental procedures that guarantee homogeneous heating. Sudden voltage drop. As mentioned in Table 2, the power supply, being analogue, may not supply the required voltage fast enough. Slow voltage reduction. Temperature recording without the power supply having been switched off immediately after the nut fell, which would lead to the electromagnet continuing to heat up during this period due to the passage of current. The immediate shutdown of power supply after nut drop. Reaction time to determine the moment of fall of the nut. Increase the sample size. Not capturing the temperature of the electromagnet in the same area. Since the heating of the electromagnet is not uniform, this would cast a shadow on the relationship between the variables. Define a specific area where the temperature will be taken in all attempts. Relationship between Temperature and the Holding Force of an Electromagnet in a Changing Magnetic Field © 2022 Global Journals 1 Year 2022 24 Global Journal of Science Frontier Research Volume XXII Issue ersion I VI ( A ) 0.47 ± 0.08 2.51 ± 0.02 0.4 ± 0.10 2.49 ± 0.01 0.36 ± 0.11 2.50 ± 0.01 0.36 ± 0.11 2.50 ± 0.01 0.39 ± 0.10 2.50 ± 0.01 supply with those of the