Lab update

Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

9-2 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

CHAPTER 9

Using Survey Research

Please click on the VOICE icons to play each slide.

9-3 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

SURVEY RESEARCH

• Different than Observational research

• no measures are administered

• Field survey

• measures are administered to assess attitudes and/or behavior

• Survey research is correlational • No causal inferences can be drawn

• Can use survey research to predict behavior

• Must ensure

• Anonymity

• Confidentiality Retrieved from http://www.rethinkdata.net/wp-

content/uploads/2016/12/Survey-2.jpg on March 9,

2017.

9-4 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

DESIGNING A QUESTIONNAIRE

• Clearly define your topic

• Yields unambiguous responses

• Operational definitions

• Focus

• Demographic information (e.g., age, gender, religion)

• Often used as predictor variables

• Target behavior or attitudes

• Often used as criterion variables

• Anxiety scale

• Honesty of news

Retrieved from https://socialanxietyinstitute.org/ on

March 9, 2017.

Retrieved from http://cdn-img.health.com March 9, 2017.

9-5 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

TYPES OF QUESTIONNAIRE ITEMS • Open-Ended

• Answer question in their own words

• Ex. Explain your views on trusting the news media.

• Drawbacks

• May not understand

• Difficult to summarize and analyze

• Restricted (closed-ended)

• Respondents are given a list of alternatives and check the desired alternative

• The news can

• A. Never be trusted.

• B. Trusted from some news agencies.

• C. Trusted from news papers.

• D. Trusted from the nightly news.

• Alternatives can be ordered or unordered

• Gives you control

http://www.teacher-certification.ca

https://www.nbrii.com/

9-6 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

• Partially Open-Ended

• An “Other” alternative is added

• Rating Scale

• Circle or check a point or number

• Number of points on the scale

• Equally spaced numbers versus a stretched rating scale

TYPES OF QUESTIONNAIRE ITEMS

Retrieved from http://www.wearerecurve.com/blog/survey-format-one-little-

change-can-lead-world-insight/ March 3, 2017.

Very Weak Very Strong

9-7 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

THREE WAYS TO LABEL A RATING SCALE • Rating scales used in experimental research and survey

research

• A Likert Scale

• five or seven -point scale used to assess attitudes

• Indicate the degree of agreement or disagreement

9-8 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

HOW TO WRITE GOOD ITEMS

• Use simple words

• Stem of a question = short and easy

• Avoid vague questions

• Don’t ask too much

• Avoid “check all that apply”

• Avoid asking for more than one thing

• Soften impact of sensitive questions Retrieved from

http://www.accountantsinlondon.info/ask-a-question

on March 13, 2017

9-9 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

ASSEMBLING YOUR QUESTIONNAIRE • Organize questions

• coherent

• visually pleasing

• Demographic items – not first

• Start with interesting question

• Question order can make a difference

• Continuity

• Keep related items together

• Sensitive or objectionable items AFTER less sensitive/objectionable items

• Logical navigational path

9-10 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

ADMINISTERING YOUR QUESTIONNAIRE

• MAIL SURVEY

• Questionnaire is mailed directly

• Convenient

• Nonresponse bias

• Unrepresentative sample

• Reduce nonresponse bias by:

• Making multiple contacts

• Phone, e-mail and Mail

• Include a small token of your appreciation

Retrieved from http://www.readexresearch.com March 13, 2017

9-11 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

INTERNET SURVEY

• E-mail

• Best for short surveys

• Web site

• Allows for more complex navigational paths

• Large samples quickly

• Bias

• Nonresponse bias

• Biased samples

• Uneven computer ownership across demographic groups

• Results comparable to other methods for most nonsensitive applications

• Controversial issues

ADMINISTERING YOUR QUESTIONNAIRE- INTERNET

Retrieved from www.ibm.com on March 13, 2017

9-12 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

• Increase response rates to Internet surveys

• Presenting all of the questions on a single web page

• Providing a progress bar

• Keep the survey as short as possible

• Use a limited number of graphic elements in the survey.

• Use radio buttons rather than drop-down

ADMINISTERING YOUR QUESTIONNAIRE- INTERNET

Retrieved from https://help.surveygizmo.com/help/radio-buttons March 13,

2017.

9-13 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

• TELEPHONE SURVEY

• Ask questions directly over phone

• Live researcher

• Interactive Voice Response (IVR) technology

• Questions must be asked carefully

• “Junk calls”

• GROUP ADMINISTRATION

• Distributed to a group at once

• a class

• Completed at the same time

• Anonymity may be a problem

ADMINISTERING YOUR QUESTIONNAIRE

Retrieved by http://catalog.sentenium.com/ on March 13, 2017

Retrieved by http://elanprojects.com.au/ March 13, 2017

9-14 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

• INTERVIEW

• Face-to-face

• Structured

• Unstructured

• Interviewer may affect responses

• MIXED-MODE SURVEY

• Using more than one survey method

• Increases response rates

• Advantages

• Increasing coverage area

• Reduces nonresponse bias

• Reduces costs

• Disadvantage

• Differential responding to different modes

ADMINISTERING YOUR QUESTIONNAIRE

Retrieved from http://www.jobinterviewtools.com/

9-15 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

ASSESSING RELIABILITY OF A QUESTIONNAIRE

• Test-Retest Reliability

• Requires multiple administrations

• Intertest interval

• Problematic if

• Ideas fluctuate with time

• Participants remember their responses

• Questions are long and boring

• Parallel Form Reliability

• Essentially the same as test-retest reliability

• Alternate form of the test is administered the second time

Retrieved from http://research-methodology.net/ March 13, 2017.

9-16 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

• Split-Half Reliability

• Assessed with one administration

• Items from one half of a test are correlated with items from the second half of a test

• Odd-even split

• Kuder-Richardson Formula (KR20)

• Assessed with one administration

• Formula estimates the average for all possible split- half reliabilities

• A KR20 of at least .75 indicates moderate reliability

• Likert Scale- Coefficiant alpha

ASSESSING RELIABILITY OF A QUESTIONNAIRE

Retrieved from https://commons.wikimedia.org/wiki/File:US-Mars-Bar-

9-17 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

INCREASING THE RELIABILITY

• Increase the number of items

• Standardize the conditions

• Score questionnaire carefully

• Clearly written questions

Retrieved from https://www.dreamstime.com/illustration/reliability.html March

13, 2017.

9-18 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

ASSESSING VALIDITY OF A QUESTIONNAIRE

• Assess validity

• Content validity

• Construct validity

• Criterion-related validity

• Concurrent validity

• Predictive validity

• Validity can be affected by a number of factors (e.g., method of

administration, unclear questions)

Retrieved from https://lunarline.com March 13, 2017.

9-19 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

ACQUIRING A SURVEY SAMPLE

• Representative sample

• Sample closely matches the characteristics of the population

• Biased sample

• Occurs when sample characteristics don’t match population

characteristics

• Produce misleading or inaccurate results

• Stem from inadequate sampling procedures

http://i.investopedia.com/dimages/graphics/representative_sample.png

9-20 Copyright © 2014 McGraw-Hill Education. All rights reserved. No reproduction or

distribution without the prior written consent of McGraw-Hill Education.

SAMPLE SIZE • Try to select an economic sample

• Includes enough respondents to ensure a valid survey and no more

• Two factors are taken into account

• Amount of acceptable sampling error

• Expected magnitude of population proportions

• Formula that is used to calculate sample size using the above

parameters (see G*Power )

• Base on similar studies

http://www.statisticshowto.com/wp-

content/uploads/2013/09/10-percent-

condition.jpg