Survey_Research_II_final - Survey Research II...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Survey Research II Dimensionality and Scaling Properties of Items Dimensionality q q Back to true score theory Assume that when you are measuring something, that you are tapping a trait that differs between members of your population in a lawful way Dimensionality One-dimensional constructs Dimensionality One-dimensional constructs Height Dimensionality One-dimensional constructs Height Shorter Taller Dimensionality One-dimensional constructs Height Shorter Taller Thirst Dimensionality One-dimensional constructs Height Shorter Taller Less thirsty Thirst More thirsty Dimensionality One-dimensional constructs Height Shorter Taller Less thirsty Thirst More thirsty Estrogen Levels Dimensionality One-dimensional constructs Height Shorter Taller Less thirsty Thirst More thirsty Less Estrogen Levels More Dimensionality Two-dimensional construct Dimensionality Two-dimensional Construct Cardiovascular System Dimensionality Two-dimensional construct Cardiovascular system Heart Dimensionality Two-dimensional construct Cardiovascular system Worse Better Heart Dimensionality Two-dimensional construct Cardiovascular System Vascular system Worse Better Heart Dimensionality Two-dimensional construct Cardiovascular system Lungs Better Worse Better Heart Worse Types of Scales One-dimensional q Easier to understand Many concepts may be one-dimensional q Multidimensional q Reflect multiple properties simultaneously Application of Scales The Major One-Dimensional Scale Types q q q Thurstone Likert Guttman What Is a Scale? The assignment... ...of objects... Lightly slapping a child's hand is an appropriate way to teach a child "no" q ...to numbers... Spanking is sometimes necessary Spare the rod and spoil the child Cumulative ...according to a rule... Discipline Discipline Construct Measure Kinds of scales q q Likert Guttman Commonalities q Four steps: 1) Create a large set of questions 2) Have judges rate those questions 3) Select a smaller set of questions based on some criteria 4) Give the smaller set of questions to your sample Likert Scaling Likert Scaling Start with a large set of items that you think all reflect the same construct. Decide how to manage multiple tasks. 20 Decide how to manage multiple tasks. 20 Manage resources effectively. 4 Manage resources effectively. 4 Decide how to manage multiple tasks. 20 Manage resources effectively. 4 Work quickly and effectively under Decide how to manage pressure multiple tasks. 49 20 Manage resources effectively. 4 Decide how to manage multiple tasks. 20 Manage resources effectively. 4 Organize the work when directions are not specific. Manage resources effectively. 39 4 Work quickly and effectively under pressure 49 Manage resources effectively. 4 Work quickly and effectively under pressure Work quickly and 49 effectively under pressure 49 Work quickly and effectively under pressure 49 Organize the work when directions are not specific. 39 Work quickly and effectively under pressure 49 Decide how to manage multiple tasks. 20 Manage resources effectively. 4 Decide how to manage multiple tasks. 20 Work quickly and effectively under pressure 49 Manage resources effectively. 4 Organize the work when Decide how to manage directions are not specific. multiple tasks. 39 20 Work quickly and effectively under pressure 49 Organize the work when directions are not specific. 39 Decide how to manage multiple tasks. quickly and Work 20 effectively under Decide how to manage pressure multiple tasks. 49 20 Manage resources effectively. 4 Manage resources effectively. 4 Work quickly and effectively under pressure 49 Organize the work when directions are not specific. 39 Decide how to manage multiple tasks. 20 Work quickly and effectively under pressure 49 Organize the work when directions are not specific. 39 Organize the work when directions are not specific. 39 Organize the work when directions are not specific. 39 Organize the work when directions are not specific. 39 Organize the work when directions are not specific. 39 Organize the work when directions are not specific. 39 Likert Scaling A group of judges rates each item on a scale of favorability where: ely . ctiv ely effe ctiv d ime ffe an et se ag n tion rce Ma ma ou for res 1 f in nt. ge 5 a eo s. na ic. ud port Ma ask 4 ecif ultit im le t t sp 2 a m hat is 3 ltip no an mu 5 2 Sc ide w are ge s c na 4 ion 3 de ma ect to 3 dir ow 5 en e 2 eh wh cid 4 1 de ecid ork D nd ew 3 na e th 54 2 atio niz ely v rm ga 4 1 ecti Or info eff 3 of ime de 55 2 et ltitu nt. ag 4 1 an mu orta M a p 3 im an 51 Sc at is 2 wh 4 3 1 3 5 2 4 1 3 2 1 ti Ra ng S et he 1 1=strongly unfavorable 2=somewhat unfavorable 3=undecided 4=somewhat favorable 5=strongly favorable 1=really bad measure of discipline 5=really good measure of discipline q q q q q Explaining right from wrong is an appropriate way to teach a child "Time out" is an appropriate strategy for teaching children right from wrong Lightly slapping a child's hand is an appropriate way to teach a child "no" Spanking is sometimes necessary I believe that a child should be seen and not heard Deciding Which Items to Select Item-total correlations q Internal consistency q q Throw out items that have a low correlation with the total (summed) score across all items. For each item, get the average rating for the top quarter of judges and the bottom quarter. Do t-test; higher t-values mean the items are better discriminators, so keep them. How Does Respondent Use? q q q q Gives a 1-to-5 rating for each item on the scale. The scale must have an odd number of options with the median being "neutral" Sum these across all items. This is why this is sometimes called a "summated" scale. 1=strongly disagree 5=strongly agree q q q q q Explaining right from wrong is an appropriate way to teach a child "Time out" is an appropriate strategy for teaching children right from wrong Lightly slapping a child's hand is an appropriate way to teach a child "no" Spanking is sometimes necessary I believe that a child should be seen and not heard Discipline Construct NO Physicial Punishment Primarily Physicial Punishment 5 25 Summated Score from Measure Questions q q q q q Explaining right from wrong is an appropriate way to teach a child "Time out" is an appropriate strategy for teaching children right from wrong Lightly slapping a child's hand is an appropriate way to teach a child "no" Spanking is sometimes necessary I believe that a child should be seen and not heard Steps q q q Frequency = what did respondents say Means = what was the average response Lowest Mean is the worst variable Explaining right from wrong is an appropriate way to teach a child VAR00001 Frequency 1 3 3 12 19 Percent 5.3 15.8 15.8 63.2 100.0 Valid Percent 5.3 15.8 15.8 63.2 100.0 Cumulative Percent 5.3 21.1 36.8 100.0 Valid 1.00 3.00 4.00 5.00 Total Time out" is an appropriate strategy for teaching children right from wrong VAR00002 Frequency 2 11 6 19 Percent 10.5 57.9 31.6 100.0 Valid Percent 10.5 57.9 31.6 100.0 Cumulative Percent 10.5 68.4 100.0 Valid 3.00 4.00 5.00 Total Lightly slapping a child's hand is an appropriate way to teach a child "no" VAR00003 Frequency 3 3 6 7 19 Percent 15.8 15.8 31.6 36.8 100.0 Valid Percent 15.8 15.8 31.6 36.8 100.0 Cumulative Percent 15.8 31.6 63.2 100.0 Valid 2.00 3.00 4.00 5.00 Total Spanking is sometimes necessary VAR00004 Frequency 5 5 5 4 Percent 26.3 26.3 26.3 21.1 Valid Percent 26.3 26.3 26.3 21.1 Cumulative Percent 26.3 52.6 78.9 100.0 Valid 2.00 3.00 4.00 5.00 I believe that a child should be seen and not heard VAR00005 Frequency 15 3 1 19 Percent 78.9 15.8 5.3 100.0 Valid Percent 78.9 15.8 5.3 100.0 Cumulative Percent 78.9 94.7 100.0 Valid 1.00 2.00 3.00 Total Means q q q q q Variable 1 = 4.31 Variable 2 = 4.21 Variable 3 = 3.89 Variable 4 = 3.42 Variable 5 = 1.26 Drop: q Variable 5 = I believe that a child should be seen and not heard Now look at "real" responses q Steps q Calculate total Variable 1+Variable2 + Variable 3 + Variable 4 + Variable 5 Examine frequency of total Does it look normal? Is there variability q Internal consistency Item-Total Correlation Cronbach's Alpha TOTAL 8 6 4 Frequency 2 Std. Dev = 2.63 Mean = 16.2 0 12.0 14.0 16.0 18.0 20.0 22.0 N = 17.00 TOTAL Internal consistency q Item total Variable 1 = -.07 Variable 2 = -.08 Variable 3 = .17 Variable 4 = .29 Variable 5 = -.27 q Cronbach's Alpha = -.02 Conclusions q q Scale REALLY bad Options Listen to judges and drop variable 5 What happens? TOTAL2 7 6 5 4 3 2 Frequency 1 0 10.0 12.0 14.0 16.0 18.0 20.0 Std. Dev = 2.69 Mean = 14.4 N = 17.00 TOTAL2 Internal Consistency q Item Total Variable 1 = .21 Variable 2 = .03 Variable 3 = .31 Variable 4 = .09 q Cronbach's Alpha = .29 Conclusions q q q Get improvement when dropping Variable 5 But still not good Options? Drop more items?? Come up with better questions?? Guttman Scaling Guttman Scaling q Also known as Scalogram analysis Cumulative scaling Guttman Scaling Start with a large set of items that you think all reflect the same construct. Decide how to manage multiple tasks. 20 Decide how to manage multiple tasks. 20 Manage resources effectively. 4 Manage resources effectively. 4 Decide how to manage multiple tasks. 20 Manage resources effectively. 4 Work quickly and effectively under Decide how to manage pressure multiple tasks. 49 20 Manage resources effectively. 4 Decide how to manage multiple tasks. 20 Manage resources effectively. 4 Organize the work when directions are not specific. Manage resources effectively. 39 4 Work quickly and effectively under pressure 49 Manage resources effectively. 4 Work quickly and effectively under pressure Work quickly and 49 effectively under pressure 49 Work quickly and effectively under pressure 49 Organize the work when directions are not specific. 39 Work quickly and effectively under pressure 49 Decide how to manage multiple tasks. 20 Manage resources effectively. 4 Decide how to manage multiple tasks. 20 Work quickly and effectively under pressure 49 Manage resources effectively. 4 Organize the work when Decide how to manage directions are not specific. multiple tasks. 39 20 Work quickly and effectively under pressure 49 Organize the work when directions are not specific. 39 Decide how to manage multiple tasks. quickly and Work 20 effectively under Decide how to manage pressure multiple tasks. 49 20 Manage resources effectively. 4 Manage resources effectively. 4 Work quickly and effectively under pressure 49 Organize the work when directions are not specific. 39 Decide how to manage multiple tasks. 20 Work quickly and effectively under pressure 49 Organize the work when directions are not specific. 39 Organize the work when directions are not specific. 39 Organize the work when directions are not specific. 39 Organize the work when directions are not specific. 39 Organize the work when directions are not specific. 39 Organize the work when directions are not specific. 39 Guttman Scaling Have many judges rate each statement with respect to whether it is favorable with respect to the construct -- a yes/no rating. s Ye ly ive . ect ely eff ctiv ime nd ffe et na se ag rce an atio M ou rm res info 1 of tant. ge e s. na ic. ud por No Ma ask ecif ultit im le t t sp 2 a m hat is ltip no an mu Sc ide w are ge No c ns na io 3 de ma ect to dir ow s en e Ye eh wh cid No cid rk de De wo nd e na 4 s e th ly atio Ye niz No tive rm ga c fo Or ffe f in ee 5 eo s tim itud . Ye ge No ult tant na Ma a m por im an 1 s Sc at is Ye No wh 3 s Ye g t in Ra Sh t ee Example: Rate Yes/No: q q q q q Explaining right from wrong is an appropriate way to teach a child "Time out" is an appropriate strategy for teaching children right from wrong Lightly slapping a child's hand is an appropriate way to teach a child "no" Spanking is sometimes necessary Spare the Rod and Spoil the child How Does Respondent Use? q q q For selected items (the final scale), they simply answer whether they agree or not (yes/no). Since each item has a scale value, we can add these values. Total score is the sum of the scale items respondent agreed to. E.g. say "yes" 5 times, get a score of 5 q There is the assumption that if the score is "X" then the respondent agreed with the first "x" items Guttman Scaling Lay out a table: Guttman Scaling Lay out a table: Respondent 7 15 3 29 19 32 41 6 14 33 Item 2 Y Y Y Y Y Y Y Y Y -Item 7 Y Y Y Y Y Y Y Y --Item 5 Y Y Y Y Y -----Item 3 Y -Y Y -Y --Y -Item 8 Y Y --------Item ... Y ---------- Guttman Scaling Lay out a table: Respondent 7 15 3 29 19 32 41 6 14 33 Item 2 Y Y Y Y Y Y Y Y Y -Item 7 Y Y Y Y Y Y Y Y --When sorted by row and column it will show whether there is a cumulative scale. Item 5 Y Y Y Y Y -----Item 3 Y -Y Y -Y --Y -Item 8 Y Y --------Item ... Y ---------- Guttman Scaling Lay out a table: Respondent 7 15 3 29 19 32 41 6 14 33 Item 2 Y Y Y Y Y Y Y Y Y -Item 7 Y Y Y Y Y Y Y Y --When sorted by row and column it will show whether there is a cumulative scale. Item 5 Y Y Y Y Y -----Item 3 Y -Y Y -Y --Y -Item 8 Y Y --------Item ... Y ---------- Note the exceptions. ...
View Full Document

This note was uploaded on 08/26/2009 for the course BB H 310W taught by Professor Saltsman,brian during the Spring '07 term at Pennsylvania State University, University Park.

Ask a homework question - tutors are online