CONTEXT OF SOCIAL WORK RESEARCH -- I

 HOW TO READ A RESEARCH ARTICLE

• QUESTION studied (or hypotheses)
• SAMPLE: who was studied
• DESIGN: experimental or not, number of groups studied, use of random assignment
• RESULTS: what the study found
• MEANING: what to make of the findings

3. Write your personal reactions to the article, whether or not it resonates with your personal or professional experience, and whether you believe it will be applicable to your area of practice & how.

Functions of research
 

• Description
• Prediction
• Explanation
• Evaluation

Place of research in social work

• Facilitate your access to information
• library research
• computer research (internet, databases)
• Emphasize the importance of systematic observation
• systematic observation is what seperates research-based practice and policy from other ways of knowing which are driven by values, politics, ethics, faith and belief
• stress the importance of systematic observation, JACSW field division requires two examples of evaluation of practice during each year in the field
• Evaluation critical under managed care current climate of accountability and cutback
• Course message: PRACTICE & RESEACH ARE INSEPARABLE; ANY SEPARATION THAT SEEMS TO EXIST IS OUT OF CONVENIENCE OR IGNORANCE

Basic social work research problem: how do you (& we & they) know that what you (we) are doing with clients & communities is effective?

How would you go about answering these questions?

• Family violence experts are very critical of homebuilders and family preservation programs, but social workers and staff of DCFS love them.  What should we do?
• Your client is getting more depressed and untalkative during sessions.  Your supervisor tells you you should let the client decide what to talk about, that you should not ask questions.  What do you do?
• The child welfare agency where you work has to cut costs, which means cutting services.  Which programs should be cut?  What are the effects of cuts?
• Prisons are overcrowded.  Should we release people on early parole? Who can safely be released?
• Does the TANF program work?  For whom does it work?  Does it work better in Chicago than elsewhere in Cook County, or the collar counties?

  Characteristics of science (which make it different from faith, belief, practice wisdom, experience, etc)

• systematic, repeated observation (record keeping)
• observers must be trained
• controlled conditions (alternative causes)
• if measures used, they must be accurate (valid) & consistent (reliable)
• objectivity is a goal (which is never realized)

   Hypotheses

• HYPOTHESIS=testable statement relating two or more variables.
• hypothesis (research) = testable statement which can be studied using research methodology
• e.g. low levels of parental support are associated with  lower self esteem, powerlessness, hopelessness, and depression (Simmons & Miller, 1987)
• hypothesis (practice) = beliefs about the way a problem is maintained and what intervention will impact it
• e.g. increasing Ms. Jones' level of support will result in improved Global Level of Functioning Scores for her son Robert
• hypothetical variables are often organized as independent and dependent
• INDEPENDENT VARIABLE (X):
• aka        intervention, policy, program, treatment, action
• DEPENDENT VARIABLE (Y or O)
• aka     outcome, indicator, change target, criterion variable, goal, objective
• In research, outcome is judged by the hypothesis
• If (head start) --> (HS grad), can test this belief by measurement
• the arrow implies causality or direction of association
• Key research issue: to what extent is the independent variable responsible for the change in the dependent variable?
• To what extent is the program responsible for the goals attained?
• To what is the therapy reposnsible for the behaviors observed?
• To what extent does the policy account for the conditions observed?
• The same variable can be independent and dependent in different hypotheses:       eg: depression
• H1: mothers who are employed outside the home are less depressed than mothers employed only in the home (depression is dependent in this hypothesis)
• H2: mothers who are depressed are less likely to seek and maintain full time jobs than mothers who are not depressed    (depression is independent in this hypotheses)

Key concerns with research:

• Is the independent variable responsible for the change in the dependent variable (issues of internal validity)? Alternately, could anything else cause the observed effects?
• Is the study generalizable to other groups (issues of external validity)? Alternately, is our sample and research design powerful enough and unbiased enough to permit us to generalize?
• Is the effect large enough that we can say it is unlikely to occur by chance  alone (issues of statistics).  Alternately, how much change is enough change?

CAUSALITY

• Causal conditions necessary:
• association between two variables
• change in 1 results in a change in the other
• can be direct or inverse
• unit of analysis : what is wrong here:
• one country is 10% catholic and 60% protestant and has a suicide rate of 5 per 1000; another country is 60% protestant and 10% catholic and has a suicide rate of 10 per 1000; Deduce: protestants are more likely to commit suicide than catholics
• one town has 60% men and 40% women and has a birth rate of 30 per 1000, and another town has 40% men and 60% women and has a birth rate of 20 per 1000.  Conclude: men are more likely to have babies than women.
• temporality can be established only by longitudinal designs
• non-spurious If an unknown or unseen or unmeasured variable which, when measured, reduces the relationship between the two variables, the relationship between the two variables is said to be spurious
• e.g. 1: violence and race are spurious (intervening variable is income)
• e.g. 2:  parent education