Logo/
Log in

Make a lecture like this from your own PDF

Drop a PDF, slides or notes here. You'll sign up, then yours starts.

0:00 / 2:45

0:00 / 2:45From Observations to Appropriate Graphs
Descriptive Statistics: From Data to Insight

Summary

Full transcript

0:00Descriptive statistics begins with a population, made up of individuals, and a characteristic measured on each individual. Its possible values are called modalities. A characteristic may be qualitative, quantitative discrete, or quantitative continuous. Once the observations are listed, count each value with its effectif, written n_i, and let the total be N. Convert counts into frequencies with f_i equals n_i over N. Each frequency lies between 0 and 1, and the frequencies sum to 1, apart from rounding. Cumulative counts and frequencies answer threshold questions: how many observations, or what percentage, are at most a given value? Finally, choose the graph to match the data. Use bars or dots for discrete values, a pie chart for categories, and a histogram for continuous values grouped into intervals. In a histogram, the rectangle's area represents the class effectif; height alone is comparable only when class widths are equal. Cumulative polygons show how observations accumulate across class boundaries.

1:09After organizing the data, describe its position and spread. The mean is a weighted average: x-bar equals the sum of n_i x_i divided by N, or equivalently the sum of x_i f_i. For the maintenance example, the mean is 6.2 interventions per day. The median requires ordered data. It divides the observations into two halves: with an odd N, take the central value; with an even N, average the two central values. The mean can be pulled by an extreme value: in the salary example, a salary of 25,000 euros produces a mean of 2728 euros, while the median is 1800 euros. Quartiles locate the lower 25 percent and lower 75 percent. For the 34-note dataset, Q1 has rank 9 and value 10, while Q3 has rank 26 and value 16. Their difference, the interquartile range, is the span containing the middle 50 percent. A box plot displays the minimum, Q1, median, Q3, and maximum. Variance measures squared distance from the mean, and standard deviation is its square root. For the maintenance data, V equals 1.7136 and sigma is approximately 1.31 interventions. Use mean and standard deviation for spread around the mean; use median and interquartile range when resistance to extreme values matters.

Ask AI about this video

Answers cite the source. Free account, no credit card.