S'W' 298

1
S.W. 298

•  
• Instructor Meekyung Han, Ph.D.
• mkhan30_at_sbcglobal.net or
• mhan_at_email.sjsu.edu

2

• Let me get to know you.
• SYLLABUS ------ ANY QUESTIONS?
• REVIEW SW 240/242

3
My Buddy

• Your buddys program focus/specialization
• Your buddys 240/242 project
• Your buddys agency and its services.
• Your buddys role there
• What is your tentative research question for SW
  298?

4
Review of SW 240

• Ethics and the researcher
• Measurement and measuring variables
• Constructing measurement instruments
• Survey research approaches
• Sampling
• Group research design
• Single system design
• Qualitative research
• Program evaluation

5
Ethics IRB proposal

• CONFIDENTIALITY, ANONIMITY
• POTENTIAL RISKS
• FREEDOM TO PARTICIPATE OR WITHDRAW
• INFORMED CONSENT

6
Measurement

• Conceptual vs. Operational definition
• Conceptual definition
• A conceptual definition is a specific theoretical
  meaning of a term, but usually NOT one used for
  describing measurement
• ex. Self-esteem is the personal judgment of
  self-worth
• Operational definition
• An operational definition is the explicit
  specification of a variable in such as way that
  is measurement is possible
• ex. self-esteem is the self-rated score of
  personal worth measured by the Rosenberg
  Self-Esteem Inventory.

7
Measurement and Measuring Variables

• 1. Variable types and levels of measurement
• 2. Measurement errors Cultural bias, social
  desirability, error of central tendency
• 3. Reliability 1) Definition 2) Types
• 4. Validity 1) Definition 2) Types
• 5. Reliability and Validity

8
Types of Variables
Continuous
Categorical
Nominal Ex)gender/ race/ occupation
Ordinal Ex) Ranked data
Interval Ex) Likert scale
Ratio Ex) age/ Height/ Weight
9
Constructing Measurement Instruments

• 1. Questions to ask before measuring
• Why do we want to make the measurement?
• Assessment and diagnosis
• Practice effectiveness
• What do we want to measure?
• Consider your focus and research statement,
  especially the variable of interest, and
  operational definitions
• In what form will the measurement be (who or what
  will do the measuring)?
• Interviews
• Questionnaires
• Self-administered questionnaires
• observations

10
Use of existing instruments vs. developing your
own!

• What are the strengths and weakness of each
  approach?
• Where can I find instruments?
• Library, reference books, internet
• Research articles where researchers use and
  publish scales
• If I cant find a good one, should I develop my
  own?
• Modify an existing one (compromise)
• Develop your own (task of pilot testing and
  establishing validity and reliability of a new
  instrument)

11
Survey research approaches

• There are three main ways in which respondents
  are asked to complete the questionnaire
• Self-administered questionnaires
• Interview surveys (face-to-face)
• Telephone surveys (or over the internet)
• Questions to be considered
• Costs (stamps, envelops, printing, etc)
• Benefits?
• Appearance presentable and interesting
• Incentives?
• Bias?
• Problems with missing data?

12
Sampling

• Probability Sampling
• Simple random sampling
• Systematic sampling
• Proportionate stratified random sampling
• Disproportionate stratified sampling
• Cluster sampling

• Non-probability Sampling
• Convenience sampling
• Purposive (or judgmental) sampling
• Snowball sampling
• Quota sampling
• Selecting informants

13

• Types of Group Research Design
• Experimental Designs
• The classic experimental research design
• The posttest only control group design
• The Solomon Four-group design
• Placebo control design
• Pre-Experimental Designs
• One-group pretest-posttest design
• One-group posttest only design
• Posttest only design with non-equivalent groups
• Quasi-Experimental Designs
• Non-equivalent control group design
• Time-series or interrupted Time series design
• Multiple time-series design
• Advanced Designs
• Factorial Designs
• Crossover Designs

14
Single System Designs?

• Why/When use Single System Design?
• Often we have no control groups, or we need to
  evaluate the effectiveness of a program or a
  group of individuals or a single person.
• Useful for immediate, inexpensive, and practice
  feedback on whether their clients are improving.
• Single systems can include a SINGLE client,
  community, organization, family, couple, setting,
  program, etc (N1 research)
• Provides a bridge between research and practice
• Types of Single System Designs
• The case study or B design
• The AB design
• The ABA and ABAB design
• The ABC and ABCD design
• Multiple baseline design

15
Why use Qualitative research?

• To investigate in-depth answers to complicated
  questions which may not be answerable through
  quantitative methods
• To compliment quantitative research in expanding
  our understanding of the factors or variables
  measured as well as discovering issues and
  factors that may not have been assessed or
  measurable
• To understand more of the processes that occur
  among variables or factors of interest
• Remember sheet comparing qualitative vs.
• quantitative research

16
Examples of Qualitative Methods

• Ethnography or Naturalistic Inquiry
• Biography
• Phenomenology
• Grounded Theory
• Content Analysis
• Case Study

17
Program Evaluation

• What is Program evaluation?
• Program evaluation is carefully collecting
  information about a program or some aspect of a
  program in order to make necessary decisions
  about that program
• The type of evaluation undertaken depends on the
  type of information you want to generate
• Types of Program evaluation
• Needs assessment
• Program implementation ( process evaluation)
• Program outcomes (outcome evaluation)
• Program impact evaluation

18
Statistics Without PainReview 242
19
Errors in Hypothesis Testing

• Type I Error reject a true null hypothesis
• Type II Error Fail to reject a false null
  hypothesis

20
The Tests

• Chi-Square test ?
• T- Test ?
• ANOVA
• Correlation
• Regression

21
Chi-Square

• Use when both your DV and IV are categorical
• The test looks for ASSOCIATIONS
• ? Use when you want to see if statistically
  significant differences exist between observed
  and the expected frequencies
• ? Use to tell if the differences in values are
  systematic or have occurred by chance
• note Cant use this test if the expected
  count in any cell is less than 5 --- its ok if
  the actual count is les tahn5, but not the
  expected count actual count and expected count
  are highly related
• Careful associated does not mean causal..

22
T-Test and ANOVAVery commonly
used test tests to compare means. Only
difference is that in T-Tests you are comparing
the mans of only 2 groups.

• ANOVA
• Use when you want to compare several means
• More than two categories of the IV, and a
  Continuous DV
• Post hoc test Need to use Scheffe or Bonferroni
  adjusted alpha level

• T-Test
• Use with categorical IVs (2 categories) and
  continuous DV
• Compare means of two groups two independent
  samples T-test and Paired samples T-test

23
Correlation

• Use when both your DV and IV are continuous
• Use when you want to see the relationships among
  variables you can use multiple variables
  together
• Can have Positive or Negative correlation
• Pearson vs. Spearman correlation coefficients
• - Mostly we will use Pearson correlation
  coefficients

24
Regression

• Use when you want to separate out the effects of
  many variables on your outcome
• you can use both categorical and continuous IVs
• You should use a CONTINUOUS DV
• Logistic Regression (Probit and Logit) used with
  continuous or categorical IVs and CATEGORICAL DV
  (We wont do these in this class)

25
Look At the Cell and Freq.

P-VALUE
Chi-square
26
When t-test stats p-value is less than .05
then, you need to use this descriptive table to
report.
T-test stats
P-value
Levens test for equal variance P is bigger
than .05 Then, report the first line t-test value
and p-value
27
When t-test stats p-value is less than .05
then, you need to use this descriptive table to
report.
T-test stats
P-value
Had negative sign post-valuegt pre-value
28
When t-test stats p-value is less than .05
then, you need to use this descriptive table to
report.
P-value
F-test stats
29
P-value you should pay attention to see which
groups are different from each other.
Post Hoc Tests