survey research #1-4

to have reliable and valid measure of a contruct

• -several items are generally better than one
• *many ways to assess the same thing

-vary the wording of items to elicit both high and low responses

-have a reason for each item

-consider the order of items

• -always include instructions
• *what do you want participant to do?

--have an organized layout

• --alwways label your scale points
• *ideally label them all

• -avoid ambiguity
• *to what extet do you approve or dissaprove how the president handles the problems
• **not specific

• -avoid double barreled questions
• *to what extent do you approve or disaprove of the President and his efforts to fix the economy?

-take care to ensure that all items measure what you think they are measuring

• -avoid leading questions
• *how positive do you feel....

-give respondents appropriate options and context for a question

-specific rating scale or highly structured recall prompts

-most reliable for concrete behaviors

• -specific contructs can be assessed in several ways
• *direct attitude assessment
• **using a scale from 1-7

• - behavioral indicators
• **on average how much do you...this a week?

-format can limit information that can be obtained

• example: do you believe....is?
• a) good for society
• b) bad for society
• *consider adding "other and specified"
• **limits the response from the individual, dont get in depth analysis of the participant

- it can be used to assess behavioral frequencies if more specific time intervals are added

-most common item type in psychological questionaire research

• -initially they were used as agree and disagree scales
• *they now got varied

• example: my mohther: 
• 5: negative , 4: OK, 3 Positive

• -ratings of bipolar adjectives in realtion to a particular domain
• *usually on a 7-point scale

• example: rate your psych 242
• easy _ _ _ _ _ _ _ _ Hard

-Clear operationalization

• -very specific and concrete
• *know what participant is responding to

-easy to quantify and use statisticalluy

- can be tested for reliability

-potential insensitivity

-often brief, simply worded, potentially superficial

• -issues or contructs are imposed on participants
• *may not exactly reflect their belifs

-general textual response: more sensitive to respondent

example: what is the most important social issue facing the country today?

-more difficult to interpret: analyze as qualitative data

-frequency counts of citations or statements

- lnkages analysis of co-occuring statements