91Ó°ÊÓ

Let the ordered sample observations be denoted by \(y_{1}, y_{2}, \ldots, y_{n}\left(y_{1}\right.\) being the smallest and \(y_{n}\) the largest). Our suggested check for normality is to plot the \(\left(\Phi^{-1}((i-.5) / n), y_{i}\right)\) pairs. Suppose we believe that the observations come from a distribution with mean 0 , and let \(w_{1}, \ldots, w_{n}\) be the ordered absolute values of the \(x_{i}^{\prime}\) s. A half-normal plot is a probability plot of the \(w_{i}^{\prime} s\). More specifically, since \(\quad P(|Z| \leq w)=P(-w \leq Z \leq w)=\) \(2 \Phi(w)-1\), a half-normal plot is a plot of the \(\left(\Phi^{-1} /\\{[(i-.5) / n+1] / 2\\}, w_{i}\right)\) pairs. The virtue of this plot is that small or large outliers in the original sample will now appear only at the upper end of the plot rather than at both ends. Construct a half-normal plot for the following sample of measurement errors, and comment: \(-3.78,-1.27,1.44\), \(-.39,12.38,-43.40,1.15,-3.96,-2.34,30.84\).

Step by step solution

01

Order the Absolute Values

First, take the absolute values of each measurement error in the list: - |-3.78| = 3.78, |-1.27| = 1.27, |1.44| = 1.44, |-0.39| = 0.39, |12.38| = 12.38, |-43.40| = 43.40, |1.15| = 1.15, |-3.96| = 3.96, |-2.34| = 2.34, |30.84| = 30.84. Next, order these absolute values from smallest to largest: 0.39, 1.15, 1.27, 1.44, 2.34, 3.78, 3.96, 12.38, 30.84, 43.40.

02

Calculate Normal Quantiles

Compute the normal quantiles using \(\frac{i-0.5}{n}\) for each observation where \(n = 10\) is the number of observations, and \(i\) is the position of the data point in the ordered list, ranging from 1 to 10. For example, for \(i = 1\)\, we calculate \(\frac{1-0.5}{10} = 0.05\). In general, the sequence will be: 0.05, 0.15, 0.25, 0.35, 0.45, 0.55, 0.65, 0.75, 0.85, 0.95.

03

Convert to Normal Deviates

For each of the quantiles found in Step 2, calculate the normal deviate using the inverse normal distribution function \(\Phi^{-1}(\cdot)\). Thus, for the sequence that was derived: - \(\Phi^{-1}(0.05)\), - \(\Phi^{-1}(0.15)\), - \(\Phi^{-1}(0.25)\), - \(\Phi^{-1}(0.35)\), ... up to \(\Phi^{-1}(0.95)\). These are the theoretical quantiles from a standard normal distribution.

04

Compute Half-Normal Quantiles

Divide each regular normal quantile from Step 3 by 2, converting them to half-normal quantiles. Remember, we are plotting against half-normal distributions. Consequently, for example, turn \(\Phi^{-1}(0.05)\) into \(\Phi^{-1}(0.05/2)\). Repeat for each quantile in Step 3.

05

Plot Half-Normal Pairs

Plot each ordered absolute error value against its corresponding half-normal quantile. The pairs will take the form \(\left(\Phi^{-1}\left(\frac{[i-0.5]/10+1}{2}\right), w_i\right)\) where \(w_i\) are the ordered absolute values from Step 1.

06

Analyze the Plot

Examine the plot for any deviations from a linear pattern. A straight line suggests normally distributed measurement errors; significant deviations may indicate outliers or that the distribution is not normal. Outliers, if present, will appear more prominently at the upper end of the plot.

Unlock Step-by-Step Solutions & Ace Your Exams!

• Full Textbook Solutions

  Get detailed explanations and key concepts

• Unlimited Al creation

  Al flashcards, explanations, exams and more...

• Ads-free access

  To over 500 millions flashcards

• Money-back guarantee

  We refund you if you fail your exam.

Over 30 million students worldwide already upgrade their learning with 91Ó°ÊÓ!

Key Concepts

These are the key concepts you need to understand to accurately answer the question.

normal distribution

A normal distribution is a fundamental concept in statistics. It describes a data set where most of the values cluster around a central mean and decrease symmetrically towards either extreme. This makes the shape of a bell curve when visualized:

• The mean, median, and mode of a normal distribution are all equal and located at the peak.
• Approximately 68% of the data falls within one standard deviation of the mean.
• About 95% lies within two standard deviations.
• Nearly all, or 99.7%, are within three standard deviations.

The normal distribution is important because many real-world phenomena approximate this shape. It is used extensively in the fields of natural and social sciences to represent random variables whose distributions are not known. Understanding this distribution is crucial as it underpins various statistical methods and tests, such as hypothesis testing and regression analysis.

probability plot

A probability plot is a useful graphical tool in statistics that helps determine if a data set follows a specified distribution, often the normal distribution. It displays data points so that if they are from the specified distribution, the plot will form a straight line.

• When constructing a probability plot, the data is ordered first.
• Quantiles from the theoretical reference distribution, like normal quantiles, are then calculated.
• Data points are plotted against these theoretical quantiles.

If the data follows the distribution being tested against (perfectly normal data against a normal distribution plot), then the points will lie approximately along a straight line. Deviations from this line suggest deviations from normality. Probability plots are especially helpful in revealing outliers and other anomalies that might not be visible in simple descriptive statistics or other summary statistics.

statistical analysis

Statistical analysis involves collecting, reviewing, and interpreting data to uncover patterns and trends. It's a vital component of research that informs decision-making across various domains like business, healthcare, and scientific inquiry.

In statistical analysis, there are several steps involved:

• Data Collection: Gathering information needed for analysis, ensuring its accuracy and relevance.
• Data Organization: Preparing the data for analysis, often involving sorting and cleaning.
• Descriptive Statistics: Summarizing the data using measures such as the mean, median, and standard deviation.
• Inferential Statistics: Drawing conclusions from the data using methods like hypothesis testing and confidence intervals.

This analysis allows us to determine the characteristics of the data, make predictions, and evaluate the relationship between different variables. Tools and methods from statistical analysis are critical for accurate data interpretation and to make informed decisions based on empirical evidence.

outlier detection

Outlier detection is a key aspect of data analysis, focusing on identifying data points that significantly differ from the rest of the dataset. Outliers can skew analysis and lead to inaccurate conclusions if not properly identified and handled.

Important aspects of outlier detection include:

• Understanding Causes: Outliers can arise from measurement errors, experimental errors, or inherent variability in the dataset.
• Methods of Detection: Visualizing data using plots such as histograms, box plots, and probability plots, which are effective in spotting outliers.
• Statistics-Based Methods: Quantitative approaches, like calculating z-scores, or using robust statistical tests, to identify and confirm outliers mathematically.

Outliers are interesting because they can indicate special cases worth investigating. Their detection is critical in ensuring data integrity and obtaining a valid interpretation of the statistical analysis as they might also suggest an opportunity for new discoveries or enhancements in data collection processes.