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

Figure 7: Scatter plot for the relationship between voltage and log [T] V. A nalysis of R esults This section delves into the possible reasons for the trends found in the collected data, as well as ways to identify the mathematical model that best adheres to the observations of the experimental results. a) Statistical analysis of linear models Population parameters are usually estimated from samples, and in order to know how much these differ from the real value, the confidence interval is used, which limits a range within which the population parameter can be located. According to the degree of confidence of a given confidence interval, the probability that the real one is present in a certain range of values can be known. Graphically, the degrees of confidence correspond to the area percentages of a normal curve, a continuous function that approximates the distribution of values of a variable. Likewise, it can be affirmed that a data set has a normal distribution if when graphing its histogram it can be plotted with a symmetric bell curve, such that the probability of approaching the real quantity is inversely proportional to the distance on the x-axis. In other words, the narrower the confidence interval, and thus the more precise the range, the smaller the discrepancy with the true value (Rio, 2021). In the present work, since it is not possible to do repetitions of the experiment for every possible temperature, it is considered that the voltages registered in the experiment are only a sample of a complete population. For this reason, it was necessary to use the Python Statsmodels library, which provides functions for the estimation of numerous statistical models including confidence and prediction intervals. Through the use of this library, the confidence intervals of Figures 8, 9 and 10 were obtained; as well as the comparison of the ordinary least squares summaries presented in Table 5. Relationship between Temperature and the Holding Force of an Electromagnet in a Changing Magnetic Field 1 Year 2022 2 © 2022 Global Journals Global Journal of Science Frontier Research Volume XXII Issue ersion I VI ( A ) Figure 8: Confidence and Prediction Intervals in Linear Regression for the Voltage vs. Temperature Plot