Using Likert scales Fabio Pittaluga FAO – Fisheries Department
Objective Introduce concept of “scaleable attribute”- or Likert scale Introduce concept of “scaleable attribute”- or Likert scale Turn abstract concepts (vulnerability) into empirically observable elements Turn abstract concepts (vulnerability) into empirically observable elements Construct set of scales to understand vulnerability (group exercise) Construct set of scales to understand vulnerability (group exercise)
Measuring livelihood aspects Social capital Social capital Leadership Leadership Conflict Conflict Mutual aid networks Mutual aid networks ….. ….. Vulnerability Vulnerability Exposure to risk factors Exposure to risk factors Capacity to respond Capacity to respond Community vs. individual responses Community vs. individual responses …. ….
Likert scales Today the food at the cafeteria was Very bad BadOKGood Very good
Types of statements EXAMPLE People know smoking damages lungs People smoke anyhow Smoking provides social distinction Attitude Cognitive Behavioral
Types of statements NoneFewSomeManyAll STATUS PROCESS Number of women with access to land for gardens Crop yields planted over the past 5 years have Substantially decreased Decreased Remained stable Increased Substantially increased
Choosing the unit of analysis INDIVIDUALS GENDER AGE ETHNICITY …… About the collectivity About the individual Statements
Attributing value to Likert scales USE SINGLE SCALE None Low SufficientGoodFluent ? NoneFew SomeMany All Linear progression Non linear progression
Direction of response NoneFewSomeManyAll AllManySomeFewNone Number of illiterate adult males in the village WRONG !!!
How many responses? STATEMENT
How many statements per variable? 1 Statement 2 Statement 3 Statement …… How much is enough?
Example VULNERABILITY
Facilitating analysis Include a self-assessment statement Very bad Bad Just right Good Very good I believe my overall condition is Include demographic variables Gender Profession Age Ethnic group
Advantages Beneficiaries participate in drafting questionnaire Individual responses point to intra-household differences Can convert qualitative observations into quantitative measures Can be filled in easily by respondent when he/she has time
Select one aspect Define aspects of vulnerability following scenarios List possible dimensions Construct Likert scales to measure dimensions Select number of statements Group exercise
Issues to consider… Do people clearly understand statements? Do people clearly understand statements? Do statements elicit logical responses? Do statements elicit logical responses? Is the number of statements sufficient? Is the number of statements sufficient? Are there too many statements? Are there too many statements? Is time necessary to fill questionnaire appropriate? Is time necessary to fill questionnaire appropriate?
Group presentations