Upgrade to remove ads
LMSW-Practice Evaluation and the Utilization of Research-5
Terms in this set (70)
The purpose of research
To discover answers to questions through the application of scientific procedures
Exploratory or formulating studies
Purpose is to gain familiarity with phenomenon or to achieve new insights into it, often in order to formulate a more precise research problem or to develop hypotheses
What does research always start with?
Purpose is to describe accurately the characteristics of a particular individual, situation, or group.
Example of descriptive study
Narrative case studies
Purpose is to determine the way in which something is associated with something else.
Studies testing causal hypotheses
Purpose is to test a hypothesis about a causal relationship between variables
A conjectural statement (prediction) of the relation between to or more variables
Guides the collection of data and also open to public scrutiny the biases of the investigators
The investigator manipulates and controls at least one independent variable and observes the way that this manipulation affects a dependent variable
Most effective method of testing a hypothesis that one variable causally influences another variable
Does not allow one to rule out in advance, with any confidence, the possibility that the effect was created by some other factor that is associated (correlated) with the presumed causal factor
The variable manipulated or controlled by the experimenter
Often denoted by x
The variable affected by x
Often represented as y
In clinical studies...
Treatment is the independent variable
Outcome is the dependent variable
Experimental group design
Consists of a comparison of control group with an experimental group
Many kinds of control groups, most common is "no treatment" group-others include wait-list, minimal contact
Receives experimental treatment
Comparison of experimental and control group before treatment and after treatment
Pre-Post (AB) Design
Comparison of some variable before and after treatment
after treatment=post treatment=B
Difficulty in ruling out alternative causes of change
Single subject designs
Allow experimental study of a single individual
Ex: Reversal Design (ABA)
Baseline-treatment-withdrawal of treatment
Does dependent variable return to baseline (pre-treatment) level?
Careful specification of the intervention and the goal to be achieved and the use of an instrument to measure states of goal attainment, a determination of the state of the problem related to the goal before the intervention is applied, and regular observations throughout the course of the change effort to see if and how the problem condition change effort to see if and how the problem condition changes
If patient were at risk for harm
Some cases it would be unethical to withdraw treatment if...
Do not delay treatment in order to obtain baseline data
Can you get the same answer repeatedly?
Are the measures obtained "true" measures of what you are measuring?
Refers to how accurate the instrument is
The extent to which differences in scores on it reflect true differences in the characteristic which it seeks to measure , or true differences in the subject being measured
Are we measuring what we think we are measuring, or something else?
The instrument is designed to capture the answers to the questions the researcher is asking
Can the results be generalized to other groups, settings, or times or from one measure to another?
Did the experimental treatments make. A difference in this specific instance?
Is the content of this measure representative of the content of the property being measured?
How well does the Beck Depression Inventory measure depression?
To what extent do certain explanatory concepts or qualities account for performance to the test?
To what degree does variation in test scores predict variation on some other measure?
Does IQ predict academic success?
Does the test focus directly on the behavior of the kind in which we are interested?
Ex: testing reading speed by timing how long it takes a child to read a passage of known length
A collection of theory and methods applied for the purpose of understanding data
Describe the data
Infer characteristics of the population from characteristics of the sample
Types of inferential statistical tests
ANOVA; t-test;chi square
Nominal (root=name) scale
Classifies objects into categories based upon some defined characteristics (sex, race, color)
Classifies without order, categories mutually exclusive
Logical ordering of categories
Ordering of cities in Illinois according to population, ordering of football teams according to number of games won, arranging students according to height
Categories have logical order no are scaled according to the amount of the particular characteristic they possess
Differences between various levels of the categories on any part of the scale reflect equal differences of the characteristic measured (equal intervals)
Ex: temperature in degrees Fahrenheit or centigrade
Equal intervals, zero is just another point on the scale and does not reflect an absolute absence of the quality being measured
Equal differences in the characteristic are represented by equal differences in the number assigned to the categories (equal intervals) and the point zero reflects an absence of the characteristic
Ex: speed in miles per hour, weight, length
Each individual within a population has an equal chance of being selected for study, and all members of the sample are selected independently of one another
Occurs when individuals selected for study are assigned to experimental or control groups according to chance
Ex: flipping a coin, table of random numbers
Most frequent score that appears in a distribution of values
The point below one-half or 50% of the scores lie
Often the best measure of central tendency in highly skewed distribution
Ex: mean size of city in Illinois, early income
Arithmetic average of scores
Most strongly affected by extreme scores
Ex: mean size of city in Illinois, yearly income
However, usually varies less from sample to sample drawn from the same population
Occurs when different experimenters administer the test differently
In sampling, occurs when one subject is more likely to be selected than another
In assignment, occurs when subjects are not equally likely to be assigned to experimental or control groups
A statement of no difference or no relationship
Ex: no relationship between the variables tested, no difference between the control group and the experimental group
Level of significance, statistical significance
Probability that difference is due to chance
P<.05 probability less than 5/100 that difference is due to chance
P<.01 probability that difference is due to chance is less than 1/100 are used in behavioral science research
Concerns human phenomena so as to better understand them as an end in itself
Conducted so as to do something better or more efficiently
Examples of Social Workers using pure research
Studies of human development and emotions, behaviors, and thought processes; group dynamics; family processes; and organizational, community, and societal conditions
Reasons for stages of research process
To determine the credibility of a piece of research, the worker must know how to evaluate each stage
Begins with a determination of the problem that the research should help to solve
Example of concern whether or not an approach is effective
Is social skills training effective in helping severely mentally ill people to function independently in the community?
Example of a question about the causes of a situation or even whether some conditions are present or absent in the situation
What conditions are more likely to be found in an agency that has poor morale than one that has good morale?
Tells the worker whether the study will be relevant or not to his or her interests
Definition of concepts
2nd part of problem formulation
Statements framed in such a way that information can be collected to support them or prove them to be false
Statement of assumptions
Conditions that are thought to be present in the situation to be investigated, but the researcher believes they will not affect the results of the research in unpredictable
Should be identified and a rational given for accepting them
Researcher does not gather evidence to test these
Researcher seek to...
Relate their problem formulation to a theory
A statement that links a number of hypotheses (that have received empirical (observations) support) with an explanation of why these findings occur as they do
Can guide us in developing and testing new hypotheses that support, modify, or negate the theory itself
Research Measurement Design-1
Decisions are made about how the data will be collected
Will subjects be interviewed face to face?
will they be asked to fill out questionnaires or does the researcher observe the subjects doing some action
Research Measurement Design-2
Decisions are made about how often and under what circumstances the research instrument(s) will be administered
Research Measurement Design-3
Decisions are made about how the collected data will be handled for confidentiality purposes, and coded and analyzed
Coding means how the data will be converted into a form that can be easily understood and reviewed and analyzed for results
Analysis consists of the decisions made about what statistical procedures will be used to determine the results and /or findings
Research Measurement Design-4
The research must relate the data to the original problem formulation so as to answer the questions posed at the point and determine whether the hypotheses have been supported or not
Data is in numerical form or where the data are converted into numerical form for statistical analysis
Provides a microscope in which to focus on specific types of information while recognizing that much has been excluded
Example of quantitative research
What is your religion? (Circle one)
1. Catholic; 2. Protestant; 3. Jewish; 4. Moslem; 5. Other
How frequently do you feel sad?
5. Always; 4. Frequently; 3. Sometimes; 2. Seldom; 1. Never
How many cigarettes do you smoke a day? ________
Seeks to be descriptive of actual events as they occur and relies on the analysis of words/texts/descriptions and not numbers for analysis
Relies on questionnaires, observations, or interviews with people
Present things as they occur in much of their complexity
Used when very little is known about the subject being studied
The researcher may wish to conduct a descriptive study in which they seek to learn more specifics
The independent variable must be specified which is often some form of intervention; the researcher must make sure that other variables that may have an effect on the outcomes are either measured so that their effects can be accounted for or that such variables are the same for all subjects in the study-an experimental control
Researcher must be sure to select the subjects randomly so that they are representative of the population from which they have chosen
Test of statistical significance difference between sample means
Test of statistical significance that measures the difference between observed frequencies and expected frequencies due to chance
Values less than .05 are statistically significant
Assessed by instrument reliability
Believed to cause or determine another variable
The impact, result, or effect
Example Taking a test preparation course increases your exam scores
Independent variable: test prep course
Dependent variable: exam scores
YOU MIGHT ALSO LIKE...
Psychology | Sdorow, Rickabaugh, Betz
Research Methods flash cards
Research Methods flash cards
Research and Statistics
OTHER SETS BY THIS CREATOR
LMSW Practice Test Questions
LMSW Practice Questions
LMSW Practice Exam_ 1
LMSW Practice Questions