Upgrade to remove ads
Only SGD 41.99/year
-
Flashcards
-
Learn
-
Test
-
Match
-
Flashcards
-
Learn
-
Test
-
Match
Gilner Morgan & Leech - Research Methods in Applied Setting
Terms in this set [75]
Specific Research Designs
Designs that describe specific types of randomized experimental and quasi experimental research approaches that help to visualize the independent variables of the study, the levels within these independent variables and when measurement of the dependent variable will take place
R
Random assignment to the group
NR
nonrandom assignment to the group
O
observation of the dependent variable
X
intervention [one level of the dependent variable]
~x
no intervention or the usual intervention [another level of the independent variable]
E
experimental or intervention of the group
C
Control or comparison group
M
matching
comparison group
group in an experimental study which receives the usual treatment or a different treatment from the intervention group and whose results will be compared to those of the intervention group to determine the effect of the intervention on the dependent variables
one-group posttest-only design
a weak experimental design where the participants are in one group, they are given a treatment and then a posttest [one-shot design]
NR E: X O
one-group posttest-only design
one-group posttest-only design example
investigator introduces curriculum [x] and then decides that it might be useful to determine whether it is working. At the end of the semester, the investigator sues some form of measurement [o] to determine the students' response to the new curriculum.
quasi-experimental designs
referred to as preexperimental designs
quasi-experimental designs
three quasi-experimental designs: one-group posttest-only design, one-group pretest-posttest design, posttest only nonequivalent groups design
one-group pretest-posttest design
a weak experimental design where all of the participants are in the same group. A pretest is conducted, then the treatment is given to all of the participants, and then the posttest is conducted.
NR E: O^1 X O^2
one-group pretest-posttest design formula
maturation
when the participants in the study change as a function of time, such as from the pre-test to posttest in the case of randomized experimental and quasi experimental research
no comparison with second group
one-group pretest-posttest design
wait-list comparison group design
all participants are administered a pretest then half of the other participants receive the treatment and the other half are assigned to the waitlist. After the treatment, all participants are administered another test, then the wait-list group receives the treatment are administered a final post-test
Danitz and Orsillo investigated the use of a 90-minute acceptance-based behavioral therapy [ABBT] session with first-year students in higher ed in hopes to decrease # of reported mental health issues. After recruiting students thru an email, all participants completed informed consent and pretest. half other participants were assigned to the workshop and half were assigned to the waitlist. The participants in the waitlist control group were provided the opportunity to attend a workshop a few months and the study was completed. Participants were given an additional informed consent from, the pretest, the workshop, and the posttest.
wait-list comparison group design
nonequivalent dependent variables
several outcome variables to the one group pretest-posttest design
NR E: X O or NR C: ~X O
posttest-only nonequivalent groups design
posttest-only nonequivalent groups design
a weak quasi-experimental design where there are two groups with no random assignment, where one group receives the treatment, then both groups complete a posttest
nonrandom assignment to groups
assignment of participants to groups [i.e., experimental vs. usual or no-treatment groups] by some process other than random assignment]
assignment bias
participants who chose or volunteered to participate in the intervention group may be different characteristics from those in the comparison group
sampling
the process of selecting part of a larger group of subjects with the intent of generalizing from the sample to the population
strengthen by adding a pretest of the dependent variable given to both groups
posttest-only nonequivalent groups design
pretest-posttest nonequivalent comparison group designs
better quasi=experimental designs
pretest-posttest nonequivalent comparison group designs
a quasi-experimental design where two groups with no random assignment are given a pretest, then one group receives a treatment and then both groups are given a posttest
NR E: O^1 X O ^2 or NR C: O^1 ~X O^2
pretest-posttest nonequivalent comparison group designs
example 1 pretest-posttest nonequivalent comparison group design
NSF and other fed agencies were interested in increasing # of teachers in math and science. Proposed that if classes at the college level were taught in an inquiry based manner, students with attitudes and interest about the subject would increase, and more students consider a career in teaching. independent variable = type of teaching with two levels inquiry based and traditional. dependent variable is attitude toward math or a math performance test. dependent variable given at the start and end of semester. in the example two college-level algebra classes at the same university.
example 2 pretest-posttest nonequivalent comparison group design
participants were from an infantry recruit training establishment in the UK where the org's senior mgmt decided that the low performing recruits should receive the treatment and the other recruits would be in the control group. researchers did not have control over who received the treatment. the researchers administered a pretest prior to the treatment of a workshop, then administered a posttest to assess the effectiveness of the workshop.
pretest-posttest quasi-experimental designs
investigator has control over the independent variable and can randomly assign the treatment to one and the traditional approach to the other. similar to random experimental design except that the participants have not been randomly assigned to groups or conditions.
very weak
issues that determine strength of quasi-experimental designs participants characteristics are not likely to be similar because no comparison group and/or no pretest [no random assignment of treatments to intact groups]
weak
issues that determine strength of quasi experimental designs participants characteristics are not likely to be similar because participants decide which group to join [self-assign to groups] [no random assignment of treatments to intact groups]
moderate
issues that determine strength of quasi experimental designs participants' characteristics are not likely to be similar because of no random assignment. Maybe if participants did not self assign to groups and no known assignment bias [no random assignment of treatments to intact groups]
strong
issues that determine strength of quasi experimental designs participants' characteristics are not likely to be similar because of no random assignment. Maybe if participants did not self assign to groups and no known assignment bias [random assignment of treatments to intact groups]
random assignment of treatments
used in quasi-experimental studies where random assignment of participants to groups is not possible; instead the treatment [experimental, usual or none] is randomly assigned to groups.
moderate-strength quasi experimental design
students scheduling classes due to availability or people choosing hospitals due to geographical convenience. the investigator is not able to randomly assign the treatment to certain groups bc the investigator takes advantage of a situation where it is known ahead of time that one group [school/hospital] is scheduled to receive intervention[or new curriculum] and the other group will not receive the intervention
time-series designs
the second general category of better quasi-experimental designs. within each type of time-series design, there are temporary treatment and continuous treatment designs
types of time-series designs
single-group time series designs and multiple-group time series-designs
single-group time series designs
quasi-experimental designs involving only one group in which a baseline is determined to be stable [via multiple pretests] prior to an intervention so that the researcher can conclude that the change in the dependent variable is due to the intervention and not other environmental factors.
single-group time-series design with continuous treatment
a quasi-experimental design involving only one group in which a baseline is determined to be stable [via multiple pretests] prior to intervention so that the researcher can conclude that the change in the dependent variable is due to the intervention that is given multiple times and not other factors
example of single-group time series design with continuous treatment
school implementing a new curriculum. observation so old curriculum might take place with standardized reading scores from previous semesters. same measurements would be examined during the new curriculum intervention. new curriculum is not a temporary intervention like a workshop but takes place continuously until replaced. design is popular when there are student records with many repeated measures that can be used for observations and when it is not possible or practical for a control group.
NR E: O^1 O^2 O^3 O^4 XO^5 XO^6 XO^7 XO^8
single-group time series design with continuous treatment
NR E: O^1 O^2 O^3 O^4 O^5 O^6 O^7 O^8
single-group time series design with temporary treatment
example single-group time series design with temporary treatment
a company that is interested in the effects of a workshop on being a team player. observations would take place prior to the workshop on some relevant measure. the workshop is given after four baseline measures, each a week apart on cooperative interactions. the workshop is a temporary intervention and observations are recorded immediately after the intervention at three later times. one would expect that if the workshop was successful, there would be an immediate increase after the intervention relative to the preceding baseline periods, and the effects might or might not be long lasting.
multiple-group time-series designs
quasi-experimental design in which a baseline is determined to be a stable prior to an intervention [via multiple pretests] so that the researcher can conclude that the change in the dependent variable is due to the intervention and not other environmental factors and there is a comparison group that receives the same number of measurements but does not receive the intervention
NR E: O^1 O^2 O^3 O^4 XO^5 O^6 O^7 O^8 or NR C: O^1 O^2 O^3 O^4 ~XO^5 O^6 O^7 O^8
multiple-group time-series design with temporary treatment
NR E: O^1 O^2 O^3 O^4 XO^5 O^6 O^7 O^8 or NR C: O^1 O^2 O^3 O^4 ~XO^5 ~XO^6 ~XO^7 ~XO^8
multiple-group time-series design with continuous treatment
random assignment of participants to groups
eliminate bias on all characteristics before the independent variable is introduced.
R E: X O or R C: X O
posttest-only control group design
posttest-only control group design
participants randomly assigned to groups [interventions or control group]then the intervention group receives the intended intervention and the control group receives a different intervention or no intervention or the usual treatment/curriculum and afterward both groups are given a posttest. groups less than 30 participants after random assignment
posttest-only comparison group design
two different interventions are used
example posttest-only comparison group design
students who signed up for college algebra were randomly assigned to one of the two classes. then one class was randomly assigned to the inquiry based condition and the other class to the traditional teaching approach. at the end of 15-week session, both classes were tested on an attitude about math inventory.
key point for posttest-only comparison group design
random assignment of participants to groups. equivalent prior to the intervention on all relevant variables and the dependent variable. differences on the dependent measure following the intervention can be assumed, can be assumed that the differences are due to the intervention and not due to differences in participant characteristics.
pretest-posttest control group design
two or more randomly assigned groups of participants are administered a pretest then the intervention groups receive the treatment then all participants are administered a post-test. participants are randomly assigned to groups, each group is pretested on the dependent variable., intervention group then receives the intervention. the control group receives traditional treatment. mortality among the control group is a threat. after the intervention period, both groups are measured again on the dependent variable [posttest]. most common randomized experimental design.
R E: O^1 X O^2 or R C: O^1 ~X O^2
pretest-posttest control group design
carryover effects
bias participants as to what to expect of the study and influence them in some way
ANCOVA [analysis of covariance]
An extension of analysis of variance that typically provides a way of statistically controlling for the effects of continuous or scale variables that you are concerned about but that are not the focal point or independent variables of the study
example of pretest-posttest control group design
pretest parents of adoptive children were given the child behavior checklist [CBCL] parent version and completed a 20 minute recorded play session. afterward the participants were randomly assigned to the treatment group, where they received child-parent relationship therapy or to a wait list control group. after 10 weeks the participants were administered the CBCL and the measurement of empathy in adult child interaction as a posttest.
solomon four-group design
experimental 2X2 factorial design where one intervention group receives the pre-test while the other intervention group does not and one no intervention group receives the pretest while the other does not; interest is in the effects of taking a pretest on the post test rather than the pretest scores themselves
R: E^1: O^1 X O^2, R E^2: X O^2, R C^1: O^1 ~X O^2, R C^2: ~X O^2
solomon four-group design; participants are randomly assigned to one of the four groups. then two of the groups [e^1 and C^1] are measured on the dependent variable [pretest]. the other 2 groups [e^2 and C^2] do not receive a pretest. then two groups [e^1 and e^2] receive the intervention. one group receives the intervention was pretested [e^1] and one group that receives the intervention was not pretested [e^2]. two groups do not receive the intervention: one that was pretested [c^1] and the was not pretested [c^2]. allows investigator to test the effects of a pretest in addition to testing effects of intervention.
solomon four-group design weakness
investigator must double the number of participants which is not worth cost and effort in most situations, so design is rarely used.
randomized experimental design with matching
participants are matched on some characteristic prior to the introduction of any of the conditions of the study, then if the independent variable has two levels, the investigator divides all of the participants into pairs of participants based on the characteristic. next investigator randomly assigns one member of each pair to the intervention group and other member of the pair to control group. all participants are administered a posttest.
M R E: X O or M R C: ~X O
randomized experimental design with matching
R Order 1 X O^1 ~X O^2, R Order 2 ~X O^1 X O^2
Within-Subjects Randomized Experimental [crossover] design
asymmetrical transfer
impact of one order [perhaps traditional treatment before the new treatment] is greater than the impact of the other order [new treatment before the traditional treatment]
how does random assignment affect bias?
random assignment should eliminate bias on all characteristics before the independent variable is introduced.
what differentiates randomized experiments from quasi-experiments?
the difference between randomized experiments from quasi experiments is that randomized experiments have random assignment to groups.
Researcher in science ed who is interested in the role of diagrams in instruction. investigate whether using diagrams in place of text will facilitate comprehension of the principles and concepts taught. develop a 12th-grade physics unit that incorporates the liberal use of diagrams. plan to compare students' knowledge of physics before and after the instructional unit. you will teach your classes using the diagram unit and other using text-only unit.
independent variable = instructional unit [active - manipulated by researcher] dependent variable = physics knowledge design name = moderate strength quasi experimental: assigning treatments to nonrandomly to groups that are probably similar
a researcher wants to study the effects of social worker support on homeless peoples' job attainment. two similar mission sites. a social worker spends a month at one of the sites, but not the other. people at the sites did not differ in average age, gender, and ed. at the end of the year she collects the following data on 2 groups from labor department records for the previous 2 years: monthly totals of # of days of employment
independent variable = presence of social worker [yes or no. Active but not manipulated by the researcher].dependent variable = number of days of employment each month for 2 years [24 data points] design name = multiple group time-series design with temporary treatment, pretty strong quasi-experimental: two groups probably similar and 12 pretest measures
health ed administering a large wellness program are interested in whether structured classes or support groups seem to have a greater influence on healthy attitudes towards food. individuals voluntarily sign up for either the classes or support groups. plan to randomly select 30 participants from classes and 30 from support groups to administer an eating attitudes instrument as a pretest and as a posttest to assess change in attitudes over time. one of the health ed expressed concern taking eating attitudes test prior to the course would affect on posttest scores bc participants will be already familiar with items on instruments and may attempt to provide a socially desirable response. what could they do to address this?
use the solomon four group design to see if the pretest makes a difference
subjects matched in pairs on key attribute variables of test scores and age and then the children in each matched pair are randomly assigned to one of the two groups, one receiving the intervention and one receiving no intervention. what specific type of design is this?
randomized experimental design with matching. design M R E: X 0 and M R C: ~X 0
Sets found in the same folderApplied Research Methods - Validity
28 terms
invadergutsy
Experiments and quasi-experiments
27 terms
Faizan_Ussenbai
Research Methods in Applied Settings
77 terms
indi45
Sampling and External Validity
29 terms
hpasheil
Other sets by this creatorAdvanced First Aid
124 terms
ritzcrackers5
Emergency Medical Responder 1
195 terms
ritzcrackers5
Emergency Responder: Advanced First Aid for Non-EM…
77 terms
ritzcrackers5
Verified questionspsychology
A child who is genetically female a. has received an X chromosome from her mother and a Y chromosome from her father. b. has received a Y chromosome only from her mother. c.. has received a Y chromosome from her mother and an X chromosome from her father. d. has received two X chromosomes from he r mother and none from her father. e. has received an X chromosome from her father and her mother.
Verified answer
psychology
Which social psychology principle influences people to perform a task better in the presence of others? a. Compliance. b. Group polarization. c. Social facilitation. d. Conformity. e. Social loafing.
Verified answer
psychology
Which of the following is true of boys compared with girls? a. Boys have a higher average intelligence score. b. Boys are better spellers than girls. c. Boys are better at detecting emotions. d. Boys are more verbally fluent. e. Boys are more likely to have extremely low intelligence scores.
Verified answer
psychology
Compared with rats raised in an enriched environment, which of the following is true of rats raised in isolation? a. Their brain cortex is less developed. b. Though neurologically similar, they fear other rats. c. Their brains have more connections. d. They have a thicker brain cortex. e. The differences between the two groups are not statistically significant.
Verified answer
Other Quizlet setshomeostasis: lesson 1
31 terms
jaskhabra
Clotting
22 terms
khreyes201
BUS010 FINAL
15 terms
a1vxis
Psych 350 exam one kahoot
20 terms
Alayna_Burch