1. Psych 230 Psychological Measurement and Statistics Pedro Wolf September 9, 2009

2. So Far Stem and leaf plots Bar plots Summarizing scores using Frequency – how a frequency distribution is created Graphing frequency distributions – bar graphs, histograms, polygons Types of distribution – normal, skewed, bimodal Relative frequency and the normal curve

3. the Normal Curve How likely is it that a certain score will occur?

4. Today…. Summarizing scores using central tendency – what is central tendency? The Mode – what it is, how to calculate it, & when to use it The Median – what it is, how to calculate it, & when to use it The Mean – what it is, how to calculate it, & when to use it – applying the mean to research

5. Range – what it is & how to calculate it Variance – what it is & how to calculate it Standard Deviation – what it is & how to calculate it Variability and the Normal Distribution Population Variance and Standard Deviation

6. Central Tendency

7. Why do we need a measure of Central Tendency? Often we would like to know the most typical or representative score of a dataset – How many drinks do students consume a week? – What are the political beliefs of students? – What is people’s favorite color? – How much do lawyers get paid? – What is the temperature in London? There are different ways to calculate a typical score. – Each way has advantages and disadvantages. Depends on: Type of data Distribution of data

8. What is a Measure of Central Tendency? Measures of central tendency answer the question: – “Are the scores generally high scores or generally low scores?” Allow us to compare values: – Average high / low temp in May in Tucson: 90ºF / 53ºF – Average high / low temp in May in St. Petersburg: 59ºF / 42ºF A statistic that indicates where the center of the distribution tends to be located

9. Measures of Central Tendency There are three commonly used measures of central tendency Mode Median Mean There is no single, perfect, measure of central tendency

10. Measures of Central Tendency

11. Example The following are the salaries of the 15 employees of a small consulting company $82,000$64,000$36,400$34,000$29,200 $29,200$29,200$28,000$26,800 $26,800 $26,800 $24,400$24,400 $24,400 $24,400 What is the typical salary of an employee in this company? How can different measures of central tendency be used to make different arguments?

12. The Mode

13. What is the Mode? The mode is the score that has the highest frequency in the data The mode is always used to describe central tendency when the scores reflect a nominal scale of measurement Can also be used for other scales of measurement Scores: 2,3,4,4,5,5,5 Mode=5

14. How to find the Mode Can find the mode by inspection (as opposed to computation) Simply the score with the highest frequency

15. Example - Mode from Raw Scores What is the mode of the following data: 141012154 141413151115 13212131413 14254711415

16. Example - Mode from Raw Scores What is the mode of the following data: 141012154 141413151115 13212131413 14254711415 Mode = 14

17. Example - Mode from Frequency Table What is the mode of the following data: Value Frequency 6565 5656 4343 3131 21211

18. Example - Mode from Frequency Table What is the mode of the following data: Value FrequencyMode = 5 6565 5656 4343 3131 21211

19. Example - Mode from Graphs Quiz #1 Scores

20. Example - Mode from Graphs Mode = 5 Quiz #1 Scores

21. Unimodal Distributions When a graph has one hump (such as on the normal curve) the distribution is called unimodal

22. Bimodal Distributions When a graph shows two scores that are tied for the most frequently occurring score, it is called bimodal.

23. Example The following are the salaries of the 15 employees of a small consulting company. $82,000$64,000$36,400$34,000$29,200 $29,200$29,200$28,000$26,800 $26,800 $26,800 $24,400$24,400 $24,400 $24,400 What is the modal salary? $24,400 – is this a good description of the typical salary?

24. Mode Advantages: – can be used with nominal data – easily identified – unaffected by extreme scores – bimodal datasets may suggest interesting subgroups Disadvantages – not necessarily a unique score – not very precise – cannot be manipulated mathematically

25. The Median

26. What is the Median? The median is the middle score of the data; the score that divides the data in half The median is the score at the 50th percentile – you did this in your homework when calculating the quartiles The median is used to summarize ordinal or highly skewed interval or ratio scores

27. How to Find the Median When data are normally distributed, the median is the same score as the mode. When data are not normally distributed, follow the following procedure: – arrange the scores from lowest to highest. – if there are an odd number of scores, the median is the score in the middle position. – if there are an even number of scores, the median is the average of the two scores in the middle. – Median score = (N+1)/2

28. Example - Median from Raw Scores What is the median of the following data: 141012154 141413151115 13212131413 14254711415

29. Example - Median from Raw Scores What is the median of the following data: 141012154 141413151115 13212131413 14254711415 First, arrange in order of magnitude 1122410 111213131313 141414141414 151515152547

30. Example - Median from Raw Scores 1122410 111213131313 141414141414 151515152547 Number of scores (N) = N=24 Median is the average of the middle two: (N+1)/2 = (24+1)/2 = 25/2 = 12.5 Average of the 12th and 13th score (13 + 14) / 2 = 13.5

31. Example - Median from Frequency Table What is the median of the following data: X fX f 6565 5656 4343 3131 21211

32. Example - Median from Frequency Table What is the median of the following data: X fN=17 6565 5656 4343 3131 21211

33. Example - Median from Frequency Table What is the median of the following data: X fN=17 6565 56Median = (N+1)/2 = (17+1)/2 = 439th score 3131 21211

34. Example - Median from Frequency Table What is the median of the following data: X fN=17 6565 56Median = (N+1)/2 = (17+1)/2 = 439th score 3131 21Median = 51

35. Median Advantages: – useful for skewed distributions – unaffected by extreme scores – useful for dividing sets of scores in to two halves (for example, high and low scorers in an exam) Disadvantages – does not take into account extreme scores – cannot be manipulated mathematically

36. The Mean

37. What is the Mean? The mean is the score located at the exact mathematical center of a distribution – the “average” The mean is used to summarize interval or ratio data in situations when the distribution is symmetrical and unimodal By far the most commonly used measure of central tendency

38. How to Find the Mean The symbol for the sample mean is The formula for the sample mean is:

39. Example Calculate the mean of the following data: 12, 15, 17, 12, 13, 9, 1, 6, 3, 12, 12, 16, 17 Mode = 12 Median = 12 N = 13  X = 12+15+17+12+13+9+1+6+3+12+12+16+17  X = 145

40. Example Calculate the mean of the following data: 12, 15, 17, 12, 13, 9, 1, 6, 3, 12, 12, 16, 17 Mode = 12 Median = 12 N = 13  X = 12+15+17+12+13+9+1+6+3+12+12+16+17  X = 145 =  X / N = 145 / 13 = 11.154

41. Example The following are the salaries of the 15 employees of a small consulting company $82,000$64,000$36,400$34,000$29,200 $29,200$29,200$28,000$26,800 $26,800 $26,800 $24,400$24,400 $24,400 $24,400 What is the mean salary? 510,000/15 = $34,000

42. Sample Mean vs. Population Mean is the sample mean. This is a sample statistic. The mean of a population is a parameter. It is symbolized by  (pronounced “mew”) is used to estimate the corresponding population mean 

43. Your Turn - Mean For the mean, we need  X and N We know that N = 18 What is  X?

44. Your Turn - Mean For the mean, we need  X and N We know that N = 18 What is  X? 10+11+12+13+13+13+13+14+14+14+14+14+14+15+15+15+15+17  X=246

45. Your Turn - Mean

46. Weighted mean The mean of a group of means Sometimes you want to compare groups with different numbers of scores Suppose you have 4 class averages: 75, 78, 72, 80. How do you find the mean? (75+78+72+80)/4 = 76.25 Only works if every class has the same number of people

47. Formula for Weighted Mean ΣfX N tot Xw =

48. Weighted mean If the Ns of the groups of scores differ…

49. Which measure to use?

50. Central Tendency - Normal Distributions On a perfect normal distribution, all three measures of central tendency are located at the same score: mean=median=mode

51. Central Tendency - Normal Distributions As the mean uses all of the information in the data, it is the preferred one to use in this case

52. Central Tendency - Skewed Distributions Use the median to summarize a highly skewed distribution

53. Which measure to use? Generally, the mean is the best measure of central tendency – All scores count in computing the mean Unless: Nominal data: use the mode Highly skewed data: use the median

54. Samples and Populations The sample mean provides a better estimate of the central tendency of a population than the sample median. The more observations in our sample, the closer the sample mean will be to the population mean. Often, the sample mean will be more accurate than any one individual

55. Applying the Mean to Research

56. Why the Mean? So, the mean is usually the best “model” of our data – The best summary of the distribution – And, we can do statistics on it – This makes it ideal for comparing groups

57. Using the mean Prediction – without knowing anything else, the mean is our best estimate Describing a score’s location – a deviation score indicates a raw score’s location and frequency Describing the population mean – ultimately we want to describe the population Summarizing Experiments

58. We compute the mean every time we have a sample of normally distributed scores The first step in an experiment is to usually to compute the mean of the dependent variable

59. Example An experiment was conducted to test the efficacy of a new diet drug. The drug was administered to rats in various dosages and the rat’s food consumption measured. The data are as follows: Rat0mg2mg4mg6mg A1210137 B1310125 C910117 D101075 E181074

60. Example An experiment was conducted to test the efficacy of a new diet drug. The drug was administered to rats in various dosages and the rat’s food consumption measured. The data are as follows: Rat0mg2mg4mg6mg A1210137 B1310125 C910117 D101075 E181074 X (0mg) = 12.4 X (2mg) = 10 X (4mg) = 10 X (6mg) = 5.6

61. Graphing Experimental Results Plot the independent variable on the X axis and the dependent variable on the Y axis Create a bar graph when the independent variable is a nominal or ordinal variable Create a line graph when the independent variable is an interval or a ratio variable

62. Bar Graphs The bar above each condition on the X axis is placed to the height on the Y axis that corresponds to the mean score for that condition

63. Line Graphs A line graph uses straight lines to connect adjacent data points

64. Today…. Understanding variability – what can it tell us Range – what it is & how to calculate it Variance – what it is & how to calculate it Standard Deviation – what it is & how to calculate it Variability and the Normal Distribution Population Variance and Standard Deviation

65. Understanding Variability

66. The mean gives us a good measure of the central tendency of our data - the average value Mean age in the class = 20.34 Does this tell us all we need to know about the distribution of ages in the class? We also need a measure of how spread out the scores are

67. Understanding Variability Measures of variability describe the extent to which scores in a distribution differ from each other The mean is our best estimate of central tendency. What is our best estimate of variability?

68. Variability - Example Same means but different variability

69. Variability - Example Same means but different variability

70. Variability - Example Three variations of the normal distribution

71. The Range

72. What is the Range? The range indicates the distance between the two most extreme scores in a distribution Range = highest score – lowest score

73. The Range - Example What is the range of the following dataset? 12, 16, 18, 23, 11, 10, 9, 4, 23, 15, 14, 13 Range = highest score – lowest score Range = 23 - 4 Range =19

74. The Range - Your turn What are the ranges of the following datasets? – 2, 6, 12, 10, 0 – 4, 7, 5, 8, 6 – 6, 6, 6, 6, 6 1) 12 - 0 = 12 2) 8 - 4 = 4 3) 6 - 6 = 0

75. The Range Though the range does give some idea of the spread of data, it is quite a crude measure Based on highest and lowest values, and so reflects the least typical scores

76. Variance and Standard Deviation

77. Most psychological research involves interval or ratio scores which approximate a normal distribution In these situations, we use two, similar, measures of variability, known as the variance and the standard deviation A measure of how different the scores are from each other

78. Variance and Standard Deviation A measure of how different the scores are from each other – calculate by measuring how much the scores differ from the mean. (Remember: the mean is our best estimate of central tendency) The variance and standard deviation indicate how much the scores are spread out around the mean Mean hours of tv watched weekly = 7.99 Are the number of hours of tv between 6 and 8 hours or between 1 and 30 hours?

79. Variance and Standard Deviation We want a measure of how much the scores are spread out around the mean Why not just take an average of the distance between each score and the mean? – Data: 4, 6, 8, 10, 12 – Mean = 8 Cannot do this, because the sum of the deviations always equals 0 – positive deviations cancel out negative deviations

80. Variance and Standard Deviation So, we want a measure which is like the average of the deviations, but which is calculated differently Conceptually, we can think of the variance and standard deviation as the typical amount that each score differs from the mean

81. The Variance One solution to the problem of deviations canceling out is to square the deviations Why? All the positive deviations will stay positive and all negative deviations will become positive - no canceling out Data: 4, 6, 8, 10, 12 Mean=8 Sum of the deviations ∑(X-X) = – [(4-8)+(6-8)+(8-8)+(10-8)+(12-8)] – [(-4)+(-2)+(0)+(2)+(4)] = 0

82. The Variance What happens when we square the deviations? Data: 4, 6, 8, 10, 12 Mean=8 Sum of the deviations ∑(X-X) 2 = [(4-8) 2 + (6-8) 2 + (8-8) 2 +(10-8) 2 + (12-8) 2 ] = [(-4) 2 +(-2) 2 +(0) 2 +(2) 2 +(4) 2 ] = [(16) +(4) +(0) +(4) +(16)] = 40

83. The Variance Does 40 reflect the average deviation from the mean? To get an average deviation, we should divide by the number of scores (N) ∑(X-X) 2 = 40=8 N 5 This statistic is known as the variance (S 2 X )

84. How to find the Variance The sample variance is the average of the squared deviations of scores around the sample mean

85. How to find the Variance The sample variance is the average of the squared deviations of scores around the sample mean Note: we can use a simpler formula to calculate the variance by hand

86. How to find the Variance - Example Estimates of professor’s age: Data: 36, 34, 45, 31, 35, 38, 36 N = 7 X = (36+34+45+31+35+38+36) / 7 = 36.43

87. How to find the Variance - Example Estimates of professor’s age: Data: 36, 34, 45, 31, 35, 38, 36 N = 7 X = (36+34+45+31+35+38+36) / 7 = 36.43 ∑(X-X) 2 = [(36-36.43) 2 + (34-36.43) 2 + (45-36.43) 2 + (31-36.43) 2 + (35-36.43) 2 + (38- 36.43) 2 + (36-36.43) 2 ] =

88. How to find the Variance - Example Estimates of professor’s age: Data: 36, 34, 45, 31, 35, 38, 36 N = 7 X = (36+34+45+31+35+38+36) / 7 = 36.43 ∑(X-X) 2 = [(36-36.43) 2 + (34-36.43) 2 + (45-36.43) 2 + (31-36.43) 2 + (35-36.43) 2 + (38- 36.43) 2 + (36-36.43) 2 ] = [(-0.43) 2 + (-2.43) 2 + (8.57) 2 + (-5.43) 2 + (-1.43) 2 + (1.57) 2 + (-0.43) 2 ] =

89. How to find the Variance - Example Estimates of professor’s age: Data: 36, 34, 45, 31, 35, 38, 36 N = 7 X = (36+34+45+31+35+38+36) / 7 = 36.43 ∑(X-X) 2 = [(36-36.43) 2 + (34-36.43) 2 + (45-36.43) 2 + (31-36.43) 2 + (35-36.43) 2 + (38- 36.43) 2 + (36-36.43) 2 ] = [(-0.43) 2 + (-2.43) 2 + (8.57) 2 + (-5.43) 2 + (-1.43) 2 + (1.57) 2 + (-0.43) 2 ] = 0.184 + 5.904 + 73.444 + 29.484 + 2.044 + 2.464 + 0.184 = 113.708

90. How to find the Variance - Example N = 7 X = 36.43 ∑(X-X) 2 = 113.708

91. How to find the Variance - Example N = 7 X = 36.43 ∑(X-X) 2 = 113.708 S 2 X = ∑(X-X) 2 = 113.708=16.244 N 7

92. How to find the Variance - Your Turn Find the variance of the following dataset: 3,3,5,9

93. How to find the Variance - Your Turn Find the variance of the following dataset: 3,3,5,9 N = 4 X = 5 ∑(X-X) 2 = [(3-5) 2 + (3-5) 2 + (5-5) 2 +(9-5) 2 ] = [(-2) 2 + (-2) 2 + (0) 2 + (4) 2 ] = [(4) +(4) +(0) + (16)] = 24 S 2 X = ∑(X-X) 2 = 24=6 N 4

94. The Variance - pros and cons Pros: The variance is a legitimate measure of variability Usefully communicates the relative variability of scores We will use it extensively in further statistics

95. The Variance - pros and cons Cons: The variance doesn’t make much sense as a measure of the “average deviation” – we have squared all of the scores, so they are unrealistically large – Professor’s Age example: Data: 36, 34, 45, 31, 35, 38, 36 Mean = 36.43 Variance = 16 (16 what? 16 squared years)

96. What is the Standard Deviation? The variance is a squared deviation score. To convert it back to the original scale, we can take the square root of the variance. This is known as the standard deviation (S X ). The standard deviation indicates the “average deviation” from the mean, the consistency in the scores, and how far scores are spread out around the mean The larger the value of S X, the more the scores are spread out around the mean, and the wider the distribution

97. How to find the Standard Deviation The sample standard deviation is the square root of the variance Note: You can use a simpler formula to calculate the variance by hand.

98. Standard Deviation - Example Find the standard deviation of the following dataset: 3,3,5,9

99. Standard Deviation - Example Find the standard deviation of the following dataset: 3,3,5,9 N = 4 X = 5 ∑(X-X) 2 = 24 [(3-5) 2 + (3-5) 2 + (5-5) 2 +(9-5) 2 ] = [(-2) 2 + (-2) 2 + (0) 2 + (4) 2 ] = [(4) +(4) +(0) + (16)] = 24 S X = √ 24=√6= 2.45 4

100. Standard Deviation - Example Estimates of professor’s age: Data: 36, 34, 45, 31, 35, 38, 36

101. Standard Deviation - Example Estimates of professor’s age: Data: 36, 34, 45, 31, 35, 38, 36 N = 7 X = 36.43 ∑(X-X) 2 = 113.708 S X = √ 113.708=√16.244=4.03 7

102. Normal Distribution and the Standard Deviation The standard deviation is a measure of how far scores are from the mean, on average

103. Normal Distribution and the Standard Deviation The standard deviation is related to the normal distribution On any normal distribution, approximately 0.34 of the scores lie between the mean and the score one standard deviation higher than the mean And, of course, 0.34 lie between the mean and the score one standard deviation lower than the mean So, 0.68 of the scores (or 68%) are within one standard deviation of the mean

104. Normal Distribution and the Standard Deviation Approximately 34% of the scores in a perfect normal distribution are between the mean and the score that is one standard deviation from the mean.

105. Normal Distribution and the Standard Deviation Why is this useful? You score 85 on an exam - is this good? If you know scores on an exam are normally distributed, that the mean score was 80 and the standard deviation was 5, what percentage of people did you score better than?

106. Normal Distribution and the Standard Deviation Why is this useful? You score 85 on an exam - is this good? If you know scores on an exam are normally distributed, that the mean score was 80 and the standard deviation was 5, what percentage of people did you score better than? You were above the mean: better than 50% You are one standard deviation above the mean: better than another 34% Therefore, you scored better than 84% of people

107. Normal Distribution and the Standard Deviation

108. Mean=66.51 Var=14.646 StdDev=3.827

109. Normal Distribution and the Standard Deviation 62.6870.38 Mean=66.51 Var=14.646 StdDev=3.827

110. Computing the Sample Variance and Standard Deviation

111. Computing the Variance and Standard Deviation The previous formulas we used for calculating variance and standard deviation showed that we were computing a measure of the average deviation of scores from the mean However, if you have to do it by hand you can use some quicker and easier formulas for calculating these statistics.

112. How to find the Variance The computing formula for the sample variance is:

113. Computing the Variance - Example Estimates of professor’s age: Data: 36, 34, 45, 31, 35, 38, 36 N = 7 ∑X 2 = 9403 (∑X) 2 = 65025

114. Computing the Variance - Example Estimates of professor’s age: Data: 36, 34, 45, 31, 35, 38, 36 N = 7 ∑X 2 = 9403 (∑X) 2 = 65025 9403 - (65025 / 7) = 9403 - 9289.286 = 113.714 = 16.244 7 7 7

115. New Terminology and Notation Review The Sum of Squared Xs – first square each raw score and then sum the squared Xs. The Sum of Xs, Squared – first sum the raw scores and then square that sum.

116. Review Find ∑X 2 for the following data: 4, 6, 7, 3, 2, 7, 2, 5 – first square each raw score and then sum the squared Xs. (4) 2 + (6) 2 + (7) 2 + (3) 2 + (2) 2 + (7) 2 + (2) 2 + (5) 2 = 16 + 36 + 49 + 9 + 4 + 49 + 4 + 25 = 192

117. Review Find (∑X) 2 for the following data: 4, 6, 7, 3, 2, 7, 2, 5 – first sum the raw scores and then square that sum. (4 + 6 + 7 + 3 + 2 + 7 + 2 + 5) 2 = (36) 2 = 1296

118. How to find the Variance (∑X) 2 = 65025 ∑X 2 =9403 N=7

119. How to find the Variance (∑X) 2 = 65025 ∑X 2 =9403 N=7 S X 2 = 9403 – 65025/7 7

120. How to find the Variance (∑X) 2 = 65025 ∑X 2 =9403 N=7 S X 2 = 9403 – 65025/7 7 S X 2 = 9403 – 9289.26 7

121. How to find the Variance (∑X) 2 = 65025 ∑X 2 =9403 N=7 S X 2 = 9403 – 65025/7 S X 2 = 113.74 7 7 S X 2 = 9403 – 9289.26 7

122. How to find the Variance (∑X) 2 = 65025 ∑X 2 =9403 N=7 S X 2 = 9403 – 65025/7 S X 2 = 113.74 7 7 S X 2 = 9403 – 9289.26 S X 2 = 16.25 7

123. How to find the Standard Deviation The computing formula for the sample standard deviation is:

124. Computing the Standard Deviation - Example Estimates of professor’s age: Data: 36, 34, 45, 31, 35, 38, 36 N = 7 ∑X 2 = 9403 (∑X) 2 = 65025 S 2 X = 16.244 S X = √16.244 = 4.03

125. Applying Variability to Research

126. Example An experiment was conducted to test the efficacy of a new diet drug. The drug was administered to rats in various dosages and the rat’s food consumption measured. The data are as follows: Rat0mg2mg4mg6mg A1210137 B1310125 C910117 D101075 E181074 X (0mg) = 12.4 X (2mg) = 10 X (4mg) = 10 X (6mg) = 5.6

127. Example An experiment was conducted to test the efficacy of a new diet drug. The drug was administered to rats in various dosages and the rat’s food consumption measured. The data are as follows: Rat0mg2mg4mg6mg A1210137 B1310125 C910117 D101075 E181074 X (0mg) = 12.4, S x = 3.13 X (2mg) = 10, S x = 0 X (4mg) = 10, S x = 2.52 X (6mg) = 5.6, S x = 1.2

128. Rat data - population distributions 02468101214 6mg, Mean = 5.8 2mg, Mean = 10 0mg, Mean = 12.4

129. Putting it all together…. Descriptive statistics - the three steps of analyzing any set of data are: 1.Consider the scale of measurement and the shape of the distribution 2.Describe where most participants scored, usually by computing the mean 3.Describe the variability of the scores, usually by computing the sample standard deviation