(5) DATA COLLECTION

issue = HOW, WHEN, FROM WHOM to collect data

how & when: todays topic, as well as a design issue

from whom = sampling & design issue

assume (before data collection method considered):

have hypotheses, OR AT LEAST . . .
have a valid research question if exploratory study
have operationalized independent & dependent variables which are valid & reliable

How, when, & from whom data collection is suggested in the research question or hypothesis

ASK CLASS FOR EXAMPLES FROM FIELD SETTINGS

1. H or RQ

2. X and Y

3. Question: how are you going to collect the data?

eg (clinical research):

Example from Single System Design

RQ What is the effect of positive parent reinforcement on the compliant behavior of an eight y/o boy?

H Consistent use of star chart (X) by Bobby's mother will significantly increase the number of days per week Bobby goes to bed on time (Y)

Practice

Perspective Behavioral

X Consistent use of star chart: mother uses star chart six or seven days per week

Y Bobby goes to bed on time: Bobby is in bed with the lights out by 9 pm and does not get up and knock at parents door

How do we actually collect the data? :

Example from Group Design

RQ what is the effect of a job training on job readiness of people on public aid?

H A job readiness program will increase the employment rate of mothers receiving TANF

Practice

Perspective Self efficacy/social learning/skill development

X Job Readiness Program: Classroom-based educational program of job-finding, interview skills + weekly support group + day care

Y Employment: Self-reported full time employment for a minimum of 60 days.

How do we actually collect the raw data?

How to collect data:

(1) observation

usually visible behavior

usually documented textually or by rating forms

preferred method of data collection in qualitative research

chapter on observation will be very useful

observation on a continuum of practitioner involvement

PARTICIPANT ______________ ACTIVE _______________ UNOBTRUSIVE

participant observer

e.g. live w/ homeless to observe mutual aid behaviors
e.g. social worker in day hospital playing softball with patients while observing socialization skills
can students do participant observation @ placement to evaluate practice?

active observer

e.g. therapist observes and rates client's level of anxiety when discussing early sexual abuse
e.g. social worker observes and documents children's visible behavior when parents argue in a family meeting
can students do active observation @ placement to evaluate practice?

unobtrusive observer

e.g. social worker observes victim-sensitive interview of a sexual assault victim behind a one-way mirror
social worker observes social skills of participants in a day hospital softball game while sitting in the bleechers
can students do unobtrusive observation @ placement to evaluate practice?

CAN WE USE SOME FORM OF OBSERVATION TO STUDY BOBBY OR EMPLOYMENT?

Bobby-yes, M will observe when he goes to bed and document
employment-probably not

(2) interview

interviews on a continuum of structure provided by practitioner

UNSTRUCTURED _____________________________STRUCTURED

unstructured interviews

e.g. have a conversation with a client to identify whether they are depressed
most interviews for most things are unstructured
can students do unstructured interviews @ placement to evaluate practice?

structured interviews

e.g. a social worker asks a series of questions about vegetative functions: sleep concentration, eating, activity
a school social worker administers the Beck Depression Inventory verbally
can students do structured interviews @ placement to evaluate practice?

CAN WE USE INTERVIEW TO STUDY BOBBY OR EMPLOYMENT?

bobby-yes, we can ask mom how he is sleeping
employment-yes, but it would be tedious

(3) testing

testing on a continuum of standardization

UNSTANDARDIZED _____________ STANDARDIZED

unstandardized

e.g. administer an opinion questionnaire to staff at an agency to identify attitudes toward certain agency policies
can students use unstandardized paper & pencil tests @ placement to evaluate practice?

standardized

e.g. adminster the Maslach Burnout Inventory to staff at an agency to identify staff burnout
can students standard paper & pencil tests @ placement to evaluate practice?

CAN WE USE PAPER & PENCIL TO STUDY BOBBY OR EMPLOYMENT?

bobby-no
employment-no

(4) existing records

records on a continuum of aggregation

INDIVIDUAL RECORDS ____________ AGGREGATE RECORDS

aggregate records

e.g. USE census data and arrest rates to measure standard of living in Illinois counties
e.g. compare delinquency rates in states which permit teacher spanking to delinquency rates in states which don't permit teacher spanking
can students use aggregate records @ placement to evaluate practice?

individual records

e.g. use case record to reconstruct client level of functioning over time
e.g. examine arrest records of men a year after being in a batterer program
can students individual existing records @ placement to evaluate practice?

content analysis

e.g. type the titles of all the articles published in six key social work journals in 1980 and 2000 the count the incidence of the word "violence"
e.g. count to ratio of 1^st person singular to 1^st person plural sentences in a clients journal
used in qualitative studies
can students use content analysis @ placement to evaluate practice?

CAN WE USE EXISTING RECORDS TO STUDY BOBBY & EMPLOYMENT?

bobby-no
employment-yes

Second issue: When to collect data (review)

Cross section: once:

e.g.

election
once-only survey
quiz 1 in 360

cross sections provide a 'snapshot' of the action
unable to predict future, identify past
since dynamics are unknown, cannot establish causal relationships

Retrospective: across time, backwards:

e.g.

looking at annual traffic death rates in a state before and after reduction of legal drunk level from 0.1% to 0.08% blood alcohol

students doing practice evaluations at JACSW often useretrospective baseline" or retrospective data collection

e.g. if target variable is increased group attendance, can go back and use past attendance as a baseline

causality can be suggested, but suggestion is weak

Prospective (longitudinal): across time, forward (election, regular survey, e.g. annual survey, daily monitoring)

comparing quiz 1 and quiz 2 is prospective
predicting college success from SAT scores
causality can be better suggested

NOTE: whether data collection is retrospective or prospective depends on when the hypothesis is made relative to the data collection

retrospective: data collected prior to hypothesis
prospective: data collected after hypothesis

Self Anchored Scale