Measuring in the Affective Domain Measuring in the Affective Domain Larry Miller and Kattlyn Wolf, OSULarry Miller and Kattlyn Wolf, OSU

For this seminar, attitudes will = perceptions, = For this seminar, attitudes will = perceptions, = opinions; = beliefsopinions; = beliefs

Thus, we will consistently use the word “attitude” Thus, we will consistently use the word “attitude” in our presentation.in our presentation.

Defining an AttitudeDefining an Attitude

““An attitude is the affect for or against a An attitude is the affect for or against a psychological object” psychological object” Louis Thurstone, 1931Louis Thurstone, 1931

evaluation ofevaluation of like or dislike oflike or dislike of positiveness or negativeness towardpositiveness or negativeness toward (Do not say “feelings”)(Do not say “feelings”)

AssumptionsAssumptions

Attitudes are predispositions to respond.Attitudes are predispositions to respond. Attitudes are persistent over time.Attitudes are persistent over time. Attitudes are susceptible to change, but not Attitudes are susceptible to change, but not

easily.easily. Attitudes produce consistency in behavior.Attitudes produce consistency in behavior. Attitudes are directional (Summers, 1970).Attitudes are directional (Summers, 1970).An attitude is like a puff of smoke that we wish to describe.An attitude is like a puff of smoke that we wish to describe.To do so, we must sample many molecules in the puff if weTo do so, we must sample many molecules in the puff if wewish to describe it adequatelywish to describe it adequately. . One molecule One molecule will notwill not do! do!

Types of Attitudinal ScalesTypes of Attitudinal Scales

Likert (summated rating)Likert (summated rating) Semantic DifferentialSemantic DifferentialWe will discuss these two, most common, scales but We will discuss these two, most common, scales but

additional information is provided in the handout additional information is provided in the handout should you have an interest in the latter two.should you have an interest in the latter two.

Guttman (cumulative)Guttman (cumulative) Thurstone (equal-appearing interval)Thurstone (equal-appearing interval)

General Criteria for Attitude StatementsGeneral Criteria for Attitude Statements Items should be a series of statements; not questions.Items should be a series of statements; not questions. Avoid statements that refer to the past rather than to the Avoid statements that refer to the past rather than to the

present.present. Avoid statements that are factual or capable of being Avoid statements that are factual or capable of being

interpreted as factual.interpreted as factual. Avoid statements with multiple interpretations.Avoid statements with multiple interpretations. Avoid statements that are irrelevant to the psychological Avoid statements that are irrelevant to the psychological

object under consideration.object under consideration. Avoid statements that are likely to be endorsed or not Avoid statements that are likely to be endorsed or not

endorsed by almost everyone.endorsed by almost everyone. Select statements that are believed to cover the entire range Select statements that are believed to cover the entire range

of the affective domain of interest.of the affective domain of interest. Keep the language of the statements simple, clear, and direct. Keep the language of the statements simple, clear, and direct.

Avoid words that may not be understood by the respondents. Avoid words that may not be understood by the respondents.

Criteria for Attitude StatementsCriteria for Attitude Statements (continued)(continued)

Statements should be short; rarely exceeding 20 words.Statements should be short; rarely exceeding 20 words. Each statement should contain only one complete thought.Each statement should contain only one complete thought. Statements containing universals such as Statements containing universals such as all, always, none,all, always, none,

and and nevernever often introduce ambiguity and should be often introduce ambiguity and should be avoided.avoided.

Words such as Words such as only, just, merelyonly, just, merely, and others of similar , and others of similar nature should be used with care and moderation.nature should be used with care and moderation.

Statements should be in the form of simple sentences Statements should be in the form of simple sentences rather than compound or complex sentences.rather than compound or complex sentences.

Avoid the use of double negatives.Avoid the use of double negatives. Statements should not contain contractions.Statements should not contain contractions. Avoid double-barreled statements: “My principal and my Avoid double-barreled statements: “My principal and my

superintendent support my efforts at innovative teaching.”superintendent support my efforts at innovative teaching.”

Likert Scales: Advantages Likert Scales: Advantages (summated (summated

rating = real name)rating = real name)Rensis Likert, Rensis Likert, 1903–19811903–1981

Easy for respondents to complete, most Easy for respondents to complete, most people familiar with the scalepeople familiar with the scale

Relatively easy to constructRelatively easy to construct Most popular attitudinal measureMost popular attitudinal measure Easy to score and analyzeEasy to score and analyze Each item considered to be of Each item considered to be of equalequal attitude attitude

value (weight) -- homogeneous itemsvalue (weight) -- homogeneous items

Likert Scale Construction Likert Scale Construction Identify the attitudinal object and delimit it quite specifically.Identify the attitudinal object and delimit it quite specifically. Compose a series of statements about the attitudinal object that are Compose a series of statements about the attitudinal object that are

half positive and half negative and are not extreme, ambiguous, or half positive and half negative and are not extreme, ambiguous, or neutral.neutral.

Establish (a minimum of ) content validity with the help of an expert Establish (a minimum of ) content validity with the help of an expert panel.panel.

Pilot test the statements to establish reliability (Cronbach’s alpha) for Pilot test the statements to establish reliability (Cronbach’s alpha) for each domain.each domain.

Eliminate statements that negatively affect internal consistency.Eliminate statements that negatively affect internal consistency. Construct the final scale by using the fewest number of items while Construct the final scale by using the fewest number of items while

still maintaining validity and reliability; create a balance of positive still maintaining validity and reliability; create a balance of positive and negative items [Remember to reverse-code when summing].and negative items [Remember to reverse-code when summing].

Administer the scale and instruct respondents to indicate their level of Administer the scale and instruct respondents to indicate their level of agreement with each statement.agreement with each statement.

Sum each respondent’s item scores to determine attitude.Sum each respondent’s item scores to determine attitude.

Likert Scale Likert Scale InstrumentInstrument Construction Construction Use the general criteria for attitude Use the general criteria for attitude

statements.statements. Begin with non-threatening, easy items Begin with non-threatening, easy items

first; demographics last.first; demographics last. Have clear instructions with an example.Have clear instructions with an example. Anticipate data entry and analysis.Anticipate data entry and analysis. Anticipate missing data on items.Anticipate missing data on items. Use approved layout techniques (see Use approved layout techniques (see

Dillman).Dillman).

Scaling of StatementsScaling of StatementsResponse scales vary. Recommend to use an Response scales vary. Recommend to use an even numbereven number of of

response categories (no neutral category) and a N/A response categories (no neutral category) and a N/A response for response for agreement agreement scalesscales

Label all response categories.Label all response categories.Since this is a summated rating scale, the scale of Since this is a summated rating scale, the scale of

measurement of the sum or mean is interval. measurement of the sum or mean is interval. Never Never analyze by item. Scale of measurement of any one item is analyze by item. Scale of measurement of any one item is ordinal.ordinal.

Anchored scalesAnchored scales: frequency, importance, etc. (Odd # = OK): frequency, importance, etc. (Odd # = OK)Pictures, thermometers, etc., may be used as scalesPictures, thermometers, etc., may be used as scalesMultiple scales per item may be used.Multiple scales per item may be used.Greater range in the scales produce more variability in the Greater range in the scales produce more variability in the

data: 8 better than 6, 6 better than 4, etc. (Correlations data: 8 better than 6, 6 better than 4, etc. (Correlations work better.)work better.)

Likert ScalingLikert Scaling

Even Number of Response CategoriesEven Number of Response Categories Label all categoriesLabel all categories Use N/A if appropriate [No neutral/undecided]Use N/A if appropriate [No neutral/undecided] Frequency, Importance, etc. [Anchored]Frequency, Importance, etc. [Anchored]

11 22 33 44

Strongly Strongly DisagreeDisagree

DisagreeDisagree AgreeAgree Strongly Strongly AgreeAgree

Costa Rica is a good location for the AIAEE conference.Costa Rica is a good location for the AIAEE conference.

11 22 33 44 55 66

Strongly Strongly DisagreeDisagree

DisagreeDisagree Slightly Slightly DisagreeDisagree

Slightly Slightly AgreeAgree

AgreeAgree Strongly Strongly AgreeAgree

Costa Rica is a good location for the Costa Rica is a good location for the AIAEE conference.AIAEE conference.

Example: Multiple Anchored ScalesExample: Multiple Anchored ScalesFor each Competency, please rate (1) the level of importance For each Competency, please rate (1) the level of importance of it to your work of it to your work andand (2) your current level of knowledge. (2) your current level of knowledge.

Level ofLevel of Level of Level ofImportance Competency KnowledgeImportance Competency Knowledge1 2 3 4 Tomato Varieties 1 2 3 41 2 3 4 Tomato Varieties 1 2 3 41 2 3 4 Tomato Wilt1 2 3 4 Tomato Wilt 1 2 3 4 1 2 3 41 2 3 4 Pruning Suckers 1 2 3 41 2 3 4 Pruning Suckers 1 2 3 4(Etc.)(Etc.)

Educational Need = (I – K) (Mean I)Educational Need = (I – K) (Mean I)Produces 3 measures: I, K and Need with 2 measuredProduces 3 measures: I, K and Need with 2 measured

Can measure more.Can measure more.

Certainty MethodCertainty Method

Why consider it? Low variability in data Why consider it? Low variability in data with typical scaleswith typical scales

Procedure: Two responses from subjects:Procedure: Two responses from subjects:(1) agree/disagree and (1) agree/disagree and

(2) level of certainty of response.(2) level of certainty of response. Semantic space between categories is Semantic space between categories is

considered and weighted.considered and weighted. Result: Greater variability in data.Result: Greater variability in data.

Certainty MethodCertainty Method

AA Item….Item…. 1 2 3 4 51 2 3 4 5

D___________________________________ D___________________________________

The The response categoriesresponse categories, then, are transformed with , then, are transformed with scores scores below for below for transformed valuestransformed values, and , and scores below that being scores below that being the "the "Certainty ScoresCertainty Scores““

D5 D4 D3 D2 D1 D/A A1 A2 A3 A4 A5D5 D4 D3 D2 D1 D/A A1 A2 A3 A4 A5 11 2 3 4 5 6 7 8 9 10 11 2 3 4 5 6 7 8 9 10 11 00 3 5 6 7 8 3 5 6 7 8 9 10 11 13 16 9 10 11 13 16(Or, other values assigned between points)(Or, other values assigned between points)

Semantic DifferentialSemantic Differential Originally created by Osgood, et al., and not Originally created by Osgood, et al., and not

specifically designed, initially, for the purpose of specifically designed, initially, for the purpose of attitude measurement.attitude measurement.

Has three primary domains (factors): Has three primary domains (factors): (1) evaluation, (2) potency, and (3) activity.(1) evaluation, (2) potency, and (3) activity.

Uses pairs of opposite adjectives in describing a Uses pairs of opposite adjectives in describing a particular object/concept.particular object/concept.

For purposes of attitude measurement, a special For purposes of attitude measurement, a special form of the semantic differential consisting of form of the semantic differential consisting of entirely adjective pairs representing the evaluation entirely adjective pairs representing the evaluation dimension may be constructed.dimension may be constructed.

Some Paired Adjectives for the Evaluation Some Paired Adjectives for the Evaluation Dimension Dimension (See Book for more)(See Book for more)

Good – BadGood – Bad Nice – AwfulNice – Awful Beautiful – UglyBeautiful – Ugly Pleasant – UnpleasantPleasant – Unpleasant Clean – DirtyClean – Dirty Sweet – SourSweet – Sour Sacred – ProfaneSacred – Profane Valuable – Worthless Valuable – Worthless

Kind – CruelKind – Cruel Honest – DishonestHonest – Dishonest Fair – UnfairFair – Unfair High – LowHigh – Low Fragrant – FoulFragrant – Foul Successful – UnsuccessfulSuccessful – Unsuccessful Reputable – DisreputableReputable – Disreputable Tasteful – DistastefulTasteful – Distasteful

(Osgood, 1957)(Osgood, 1957)

Semantic Differential ProcedureSemantic Differential Procedure Identify the concept to be measuredIdentify the concept to be measured Generate a list of approximately 7 or 8 bipolar Generate a list of approximately 7 or 8 bipolar

adjectives with an number of positions between adjectives with an number of positions between each pair. (Subjects lose focus after 8)each pair. (Subjects lose focus after 8)

Administer the scale and instruct respondents to Administer the scale and instruct respondents to identify where, on the continuum between the two identify where, on the continuum between the two adjectives, their beliefs about the concept lie.adjectives, their beliefs about the concept lie.

The spaces or positions between the adjectives The spaces or positions between the adjectives become categories with a numerical value (e.g. become categories with a numerical value (e.g. 1=unfavorable and 6=favorable) and responses are 1=unfavorable and 6=favorable) and responses are summed to determine attitude.summed to determine attitude.

GRADUATE SCHOOL GRADUATE SCHOOL ADMISSION ADMISSION (Concept)(Concept)

(Adjective Pairs with scaling)(Adjective Pairs with scaling)

Rough___: ___: ___: ___: ___: ___SmoothRough___: ___: ___: ___: ___: ___Smooth

Fair ___: ___:Fair ___: ___: ___: ___: ___: ___ Unfair___: ___: ___: ___ Unfair

(Add other sets of bipolar adjectives as (Add other sets of bipolar adjectives as desired)desired)

StrengthsStrengths Easy to constructEasy to construct Short and quick to Short and quick to

administeradminister Usually highly reliableUsually highly reliable Correlate highly with Correlate highly with

scores from Likert or scores from Likert or Thurstone scales.Thurstone scales.

WeaknessesWeaknesses Long directions for Long directions for

administration can be administration can be a problema problem

Blatant and Blatant and transparent in their transparent in their purpose and can invite purpose and can invite bias or socially bias or socially acceptable responsesacceptable responses

May fail to isolate the May fail to isolate the evaluative dimension.evaluative dimension.