Research Methods Final

Front 1

chapter 19 methods

Back 1

called styles

Front 2

the harvard method

Back 2

the most widely used style of cting for social and behavioral sciences

Front 3

apa

Back 3

the american psych associaton
-adopted and modified the harvard method
-the publication manual of american psych associations describes how to use method

Front 4

in the text of a lit review

Back 4

-the last name of authors of a reference can be mad the subj. of the sentence
- good when a reviewer compart and contrast findings or thoughts of two or more authors
-good when they make a common pt
-except in school it is bad to string references togehter for a single pt -instead should use eg

Front 5

the reference list at the end of the research report should include

Back 5

- references for only those citations that were cited in the body of the report
-listed alphabeltically by authors last names

Front 6

for journal articles in apa

Back 6

-the title in lowercase except for first word and first word after colon
-title of journal is italisized followed by the volume which is also italicized

Front 7

material found on web apa

Back 7

-date which material accessed

Front 8

from book apa

Back 8

-title of book lowercase except for first word and first word after colon and is italicized

Front 9

conceptual def

Back 9

-dictionary def
-when a resarcher wants to com topic of his research to someone else

Front 10

operationalizing

Back 10

-redefining a variable in terms of physical steps
-no attempt in operationalizing a variable will result in complelety operationalized def because there r an infinate number of physical characteristics that might b addressed in any def
-so instead of looking for a completely operational def look for good enought opermit replicaiton

Front 11

operational def

Back 11

-wen researchers operationalize a veriable they create this

Front 12

replication

Back 12

attempt to confirm results of study by conducting it again in the same way
-there is subj in applying this criterion of adequacy

Front 13

highly operational def

Back 13

-not necesarily relevant or meaninglful

Front 14

treatments

Back 14

-given in experiments should b operationalized

Front 15

if researvher fails to give operational def of all key varialbes

Back 15

-report is vague, results r often limited in usefullness

Front 16

sample

Back 16

-is a group in which researhcers r ultimately interested
-after drawing sample study it and then make inferencs for the whole pop

Front 17

census

Back 17

-meaures every member of a pop

Front 18

poor sample

Back 18

-lead to incorrect inferences

Front 19

unbiased sample selected by

Back 19

-giving every member of the pop pan equal chance of being inluded in the smaple
-one way out of a hat
-result =simple random sample

Front 20

bias in sampling can be caused by

Back 20

-if dont id all members of pop

Front 21

collect bias sampling when they

Back 21

-use samples of convenience(accidental sampling)
-volunteerism-two forms
-simply call for volunteers
-even if id random sample from pop

Front 22

simple random sampling

Back 22

-every member of a popis given an equal chance of being included in the sample
- ex. of a random sample is a hat
-or a table of random numbers

Front 23

table of random numbers

Back 23

-numbers that r unrelated to eachtoher or anything else
-to use must use must give every member of pop a number name with same number of digits

Front 24

error created by random sampling

Back 24

-sampling error
-if use adequate size sampling errors go down
-nonchance errors(not reduced by increased sample size)
-

Front 25

systematic sampling

Back 25

--equivalent to simple random sampling
-every nth participant is seleccted ncan b any number
-potential probs-pop could b ordred in a certain way that makes it not random
-should go completely through the alphabet

Front 26

precision

Back 26

-discussing the magnitude of samping errors
-results r more precise when researchders redue samping errors

Front 27

the two major ways to reduce sampling erros

Back 27

-increase sample size,use stratificaiton in conjunction with random sampling(stratified random sampling)

Front 28

to obtain a stratified random sample

Back 28

-first divide pop into strata ex. men and women then draws seperately from each strata
-same percentage not number of participants from each strata
-sometimes percentage may not b rep of entire group
-

Front 29

increase precision by

Back 29

-using mutliple strata in selecting a given sample
-having larger number of varialbes in a strata is good as long as the varaiable are independent and relevant

Front 30

primary purpose of strat

Back 30

to ensure that diff subgroups r rep in correct porportions

Front 31

topic27
measure

Back 31

-generic term for any type of measurement device(test, questionnaire,interview schedule, personality scale)
-validity is addressed in this section
-also called instrument(section would b titled instrumentation)

Front 32

valid

Back 32

-determined by the extent to which it measures wat it is designed to measure and accureatly performs the function it is supposed to

Front 33

validity

Back 33

-relative to the purpose of testing
-is a matter of degree
- so discuss how valid something is not if it is valid

Front 34

perfect validity is elusive

Back 34

-almost all tests tap only a sample of the behavior underlying the constructs being measured
-some traits researchers want to measure r inherently difficult to measure
-quantitative researchers try to reduce behaviors to numbers while qualitative yield words to describe the extent to which traits r present

Front 35

content validity

Back 35

--to determine this of a measure make judgments on the appropriateness of the content

Front 36

3 principles for writing achievement tests with high content validity

Back 36

-broad sample of content
-important material emphasized
-questions should b written to measure ethe appropriaten skills

Front 37

content validity used to evaluate other meausres

Back 37

-self-concept
-should consider within it physical self concept, academic self concept, social self concept
-based on the research being conducted one might be emphasized with more questions
-want common shared meanings from examinees, so shuld consider ed levels

Front 38

face validity

Back 38

-judgments r made on whether a measure appears to be valid on the face of it
-primarily a pr concern
-prefer to use tests that look like they r related to their purpose because promotes public acceptance of testing and motivates examinees to do their best

Front 39

a test can have high face validity for a certain purpose while having

Back 39

low content validity
-researchers occasionally use this
-ex. vaigra when researchers want to disguise the true pupose of the research

Front 40

empirical approach to research validity

Back 40

-make planned comparisons to see if a measure eyelids scorses that relate to a criterion

Front 41

criterion

Back 41

-the set of ratings( the standard by which the test is being judged)

Front 42

predictive validity

Back 42

-to wat extent does the test predict the outcome it is supposed to predict
-get more info if look at rand order of individuals on the test in comparison to their rank order on the ratings

Front 43

a way to look at a tests predictive validity

Back 43

-is to compute the validity coefficient

Front 44

validity coefficient

Back 44

-a correlation coefficient used to express validity
-coefficients range from 0.00 validity to 1.00 perfect validity

Front 45

0.00

Back 45

there is no relationship between the ranks on the test and the ranks on the criterion

Front 46

concurrent validity coefficient

Back 46

-obtained by admin the test and collecting the criterion data at the same time

Front 47

criterion related validity

Back 47

-predicitvie validity coefficient
-concurrent validity coefficient

Front 48

employment test validated against supervisors raitngs

Back 48

-researcher surprised to obtain a coefficient greater than .60 and not surpsrsied by a .20 or less
because success on the hjob is a complex construct involving many traits such as interpersonal skills, psychomotor skils. work habits and more(so one test especially a paper and pencil test cant be expected to accurealty predict all traits
-coefficnetns r not higher because criteria such as supervisor ratings r less than perfectly reliable and valid

Front 49

some researcher determine the empirical validity of a tes tthat is not desigened to predict future behavior

Back 49

- ex.ASI (addiction severity index) was validated by correlating scores on it with scores obtained using the expensive and time consuming original version
-or could correlated scores on a test that is being validated with scores on a diff test

Front 50

construct validity

Back 50

-relies on subj. judments and empricial data

Front 51

construct

Back 51

-a collection of related behciors that r associated ina meaningful way
-construct does not have a physical being outside of indicators(ex. lethargy with depression)
-infer its existence by observing the collection of related indiciators
-any one indicator can b related to several constructs

Front 52

to determine construct validity of a measure begin by

Back 52

-hypothesizing about hw the construct the measure is designed to mesur should affect or relate to other variables

Front 53

if the researcher finds no relationship between the scores obtained on the depression scale and success in college this means

Back 53

either
-the scale lacks validity for measureing depression
-the hypothesis is wrong

Front 54

if the researcher finds a relationship with scores obtained on the rnew depression scale and success in college this means either

Back 54

-the depression scales is to some degree valid for measuring depression or the depression scale measures a variable other than depression that is also related to grades earned in college

Front 55

determining construct validity is

Back 55

complex matter involving judgment and empirical data
-offers only indirect evidence regarding validity of measure
-for direct need creiterion related validity

Front 56

a test is reliable if

Back 56

-if it yields consistent results

Front 57

a test with high reliability may have

Back 57

low validity

Front 58

when evaluating measures

Back 58

validity is more important than reliability

Front 59

interobserver reliablity

Back 59

--test hypothesis that tall customers r waited on more quickly than short
-1-classify as tall or short,2-count the number of secs from the time each customer eneters a store to the time a salesperson greets them
-to increase validity have two ppl independently observe \
-then determine the percentage of participants that were put in the same category by both obserers
-to comptye relationship of seconds between observers could use coreelation coefficient

Front 60

when use correlation coefficient to describe reliability

Back 60

called reliability coefficients

Front 61

when use reliability coeffiecients to describe the agreement between observers

Back 61

caked interobserver reliability coefficient

Front 62

interobserver reliability-

Back 62

usually obtain measures at the same time

Front 63

test retest reliability

Back 63

measure at two diff pts in time

Front 64

parallel reliability should b determined

Back 64

-wen published test in two parallel or equivalent forms that r designed to b interchangeable -diff items that cover the same content

Front 65

how high should a reliability coefficient be

Back 65

.80 or higher but when group averages based on group particiapnts of 25 or more .5 is ok because deermines the reliability of individueals score

Front 66

two methods for determining the reliability of a test or scale

Back 66

- test-retset reliability and parallel forms of reliability

Front 67

split half reliability

Back 67

-use sores frm a single admin of a test to examine the consistency of test scores
-admin a tst but scores the items in the test as though they consisted of two separate tests
-to do this susal ly do odd-even split- examine just the odd then just the eventhen correlates yielding the split half reliability coefficient-checks consistency of scores with in the test itself thus belongs to internal consistencey

Front 68

alt to split half methods for estimating internal consistency

Back 68

cronbachs alpha

Front 69

cronbachs alpha

Back 69

symbol is the x with the loop on the left
-single admin of test
-then math procedures r used to obtain the equivalent of the averagel of all possible split half reliability coefficents
-used more than split half because easier and on computer

Front 70

internal consistency measures

Back 70

-consistency amond the items within a tes tat a single pt in time

Front 71

parallel form sreliablity

Back 71

-consistency of scores over time

Front 72

first element in a research paper

Back 72

title
-10-15 words
-should name all the major variables
-pop of special interest mentioned
-no results
-title (under authors name and affiliation), then abstract(put intro/ lit review here), then method(subheading, particpant

Front 73

abstract

Back 73

-summary of research
-the purpose is to give readers an overview which helps them determine whether the research is relevant to them

Front 74

intro

Back 74

-follow abstract
-in short reports like journal articles and term projects the intro contains a lit review
-no main heading

Front 75

methods -participants

Back 75

-the number of them, how they were selected, and demo such as age, and gender

Front 76

methods-measures

Back 76

-describes the measure that were used
-quantitative researchers use highly sturctered mesaures that can be obj scored
-qualitative use less structred measures often semi structure interviews

Front 77

optional subheading under methods

Back 77

-procedure
-physical steps taken by researchers to conduct the research

Front 78

methods-data analysis

Back 78

-the metods used to analyze the data is described \
-optional in quantitative wen standard stats techs r used but needed for describing the type ogf qualitativbe research used eg describing how grounded theory or consensual qualitative reserach was implemented to analyze the data

Front 79

results

Back 79

-76

Front 80

discussion

Back 80

-present their interpretation of the outcomes of their research

Front 81

references

Back 81

should conclude report

Front 82

body of typical research report

Back 82

begins after abstract with no heading

Front 83

intro comments

Back 83

-r ususally integreated with the lit review
-most researchers cite lit in the first paragraph

Front 84

1st purpose of a integrated intro and lit reveiew

Back 84

-to id the research and prob area
-2nd is to establish the importane

Front 85

first paragraph of intro

Back 85

-should b reasonably specific to the research prob

Front 86

one way to establish the importance of a prob

Back 86

-is to cite stats indicating its prevelance
-describe the severity of its impact on individuals

Front 87

lit review

Back 87

- should not consist of a series of summaries
-generalizations of findings should b stated with references supporting them group e dtogether
-to provide depth it is ok to discuss selected studies in detail

Front 88

most research is designed to

Back 88

fil in gaps in knowledge about a particular topic

Front 89

a combo intro/lit review should end

Back 89

-witha statement of the specific research hypothesis, purpose,and or question
-wen several should b numbered or lettered

Front 90

participants

Back 90

-number of participants should b stated
-should contain info on demo
-metion method of recruitment
-respnse rate if known
-wen sample of convience is used
-if informed consent was obtained

Front 91

demo

Back 91

-a wide variety may b reported
-some demos r more relevant to a topic
-when large number could b put in a stats table
-it is important if two or more distict groups were compred in a stuidy

Front 92

quantitative researhcers emphasize

Back 92

-random selection of participants from pops
-so if random selection is used in quantitatvie study it is important to mention that

Front 93

qualitative researchers

Back 93

-employ purposive samples -wen u deliberatley select individuals who they think will b a rich source of info on a particular reserch topic
-should explain their reasoning and steps they took
-without info ofthis tpe reader sr likely to assume tha ta somple of convenience was used
-if saturation was used o determine sample size should b indicated

Front 94

limitations

Back 94

-can b mentioned under participants usually mentioned under discussion

Front 95

quantitative researhcers

Back 95

-obj measures such as mulitple choice tests and attitude scales(eg strongly agree to strongly disgree)
-wen used previous published measure
will cite and reference it
-specialized measures r published only in the sense that hey were included ina journal article with these the origianal source should b cited

Front 96

wen describing a measure in quantitative research

Back 96

1. name th etrait the measure is desinge dto measure
2. provide a sample item
3. indicate the reliability and validity
-wen space permits can give more detail

Front 97

when measure has been widely used

Back 97

-should b indicated
-wehn they create new measures because info on them is not availbale elsewhere need detailed info on them in their reports

Front 98

qualitative researchers

Back 98

-ususally use less sturcterd measure than quantitative
-should explicitly describe specific info ontheir measuredment methods
-most widely used is semi structured interview

Front 99

when describing measure in qualitative research

Back 99

1-where interview was conducted and who with
2-focus of the questions
3-provide sample question and prompts
4-how long interview lasted
if you have space give background on interviwer

Front 100

in qualititave interview

Back 100

-focus groups used
-should indicate how many and how many participants in each focus group

Front 101

results

Back 101

-for more than one hypothesis, purpose or question
-state in seperatly in order they were presented

Front 102

in quantitative research

Back 102

-ussusally dont indicate how sats results were computed eg comp program used
-no reference for standard stat formula
- inthese reports dscriptive stats should b reported first followed by rsults of test of stat sig

Front 103

presnet in table wen

Back 103

-number of related stats to a report

Front 104

results section for qualitative research

Back 104

-usually a discusion of the steps taken to analyze the data ussually under data analysis
-here the method of analysis should b named and the steps taken to implement the method describwed
-usually few if any stats reported instead discuss the primary thems that emerged from the data and any theories develped during th eanalyusis

Front 105

discussion section

Back 105

-begin by referecning the original purpose of the research which may have been stated as a hypothesis, question, or purpose
-then indicate to the extent that the purpose was achieved as reped in the results
- if the effect size is small or methodological weaknees dont say the hyp was strongly supported

Front 106

next in the discussion section

Back 106

-compare and contrast the results reporeted in teh lit review
-if any of the results of the curren tstudy r surprising in light of the info in the lit reciew should b pointed out
-wen resulted have practical implications these should b stated and wat actions a particular person, org, or group should perform

Front 107

limitations

Back 107

-in discussion section
-how they might effect interpretation
-typically in sampling and measurements a variety of other methodological probs ma have affected results

Front 108

future research

Back 108

-in discussion section
-may b specific because it is of so lil use to consumers of research to simply state tha tmore research is neede d

Front 109

if discussion section is long

Back 109

-provide subheadings
-such as research purpsoe and results, implications, limitations, directions for future research

Front 110

purpose of an experiment

Back 110

-to explore the cause and effect relationships

Front 111

classic design for exploring the cause and effect relationship

Back 111

-pretest postest radnomized control group design

Front 112

-pretest postest radnomized control group design

Back 112

-assigning p at random to groups so no biases in assgn
-two goups of p r equal besides random diffs
-experimental groups makes greater gais from pretest to postest than the control gorup
and if the two group shave been treated the same in al respects except for the experimentsal treaatment

Front 113

the diff is atributed to two causes

Back 113

-treatment
-random errors

Front 114

random errors

Back 114

-can use inferential stats to asse sthe role of random errrors in creating any observed diffs
-can tolerate because they results can b interpreted in light of their possible effects
-there is no generalizable statistical tech for assesing the influence of bias so researchers prefer to avoid biases in the assign to the two conditions by using random assignment

Front 115

r

Back 115

random assignment to groups

Front 116

o

Back 116

obervation or measurement

Front 117

x

Back 117

experimental treatment

Front 118

control condition

Back 118

no symbol so repped by blank
-may include some type of treatment so it is the standard for the effectivenes sof the experimental treatment

Front 119

advantage of using the pretest

Back 119

-is that it permits researchers to determine how much each group has gained

Front 120

prob caused by pretest

Back 120

-it can sensitize participants to the experimental treatment and become a part of the treatment
-called pretest sensitization
-or reactive effect of testing
-can b overcome by conducting an experiment without a pretest which is the postest only randomized control group

Front 121

combo of the pretes tand treatment

Back 121

changes observed in the ecperimental group may be the result of this

Front 122

solomon randomized four group design

Back 122

has four rows of symbols and four groups first two rows same as deisng 1 and second two design 2
-two experimetns conducted at the same time

Front 123

adv of design three

Back 123

-researchers can compare the first two gorup s to determine how much is gained is achieved and can slo comaore the last two groups to determine whetehr the the treatment is more effective than the control condition in the absence of a pretest

Front 124

drawback of design 3

Back 124

-must begin witha reasonable large pool of participants so that when they r divide dinto four groups each of the groups wil have a sufficent number to yield reliable results
-if have less than 48 bad idea

Front 125

true experimental designs

Back 125

-characterized by random assignment no treatment
-all three designs are this

Front 126

threats to internal validity

Back 126

-major chaanges to a p behavior in an experiment is it reasonable to attribute these changes to the treatment
-there may b explanation to these change in other than the treatment

Front 127

internal threat

Back 127

history
-maturation
-instrumentation (changes in measurement used from the pretest to the postest)
-testing-the effects of the pretest on the performacne exhivited on the postest
-stat regression

Front 128

statistical regression

Back 128

-occurs only if participacnt r selected on teh basis of their extreme scores

Front 129

threat to internal valididty

Back 129

-wen researchers use tow comparison groups that r not formed at random (intact group or previously exisitng group is shown in the diagrams with X and O as a line through the middle)

Front 130

seleciton

Back 130

-wen researchers do not assign p to the two goroups at random there is a ver ystorng possibility that two groups r not initially the same in all imporant respects which is a threat
-can intereact with other thteats to internal valitdity
one of which is called selction history interaciton

Front 131

selection history interaction

Back 131

-the seleciton of ps for the two groups was not at random the may b systematically subjected to diff life experiences

Front 132

selection maturation interaction

Back 132

-the two goruop s on average were at sometwhat diff developmental stages at the time of pretest which lead to diff rates of maturation in the two groups wchi hcould affect th eself concept

Front 133

selection can interact with threat called morality

Back 133

--wchih is differential loss of p fro the groups to b comparted

Front 134

all threates to internal validity can b overcome by

Back 134

true experimental design whre p r assigned at random to experimental and control conditions
-because random assignment has no bias both th eexpeirmental and control group r equally as likley to experience the same environmental events mature at the same rates and drop oout at the same rates and so on

Front 135

threats to external validity

Back 135

-seleciton bias
-reactive effects of testing (or pretest sensitization)
-multiple treatment interference

Front 136

selection bias

Back 136

-its interaciton with experimental (indepedent) variable
-if a sample is biased it is hard to generalize it to a pop
-no generalizations should b done in this case

Front 137

reactive effect of testing

Back 137

-the possibility that the pretest might ingluence how th ep respond to teh experimental treatment

Front 138

muitiple treatment interference

Back 138

-occures wen a goroup of p is givven more than one treatment

Front 139

external validity

Back 139

-to who m and under wat circumstance can the results b generalized

Front 140

internal validity

Back 140

-is the treatment in this particular case responsible for the observed changes

Front 141

preexperimental designs

Back 141

-designs r of limited value for investigation casue and effect relationships becasue of their poor internal validity
-one-grou pretest postest design
-one shot case study
-static group comparison design
-useful in prelimary pilot studies in which a researchers goal is to try out potential treamtents and measuring tools to learn more about their acceptability and accuracy

Front 142

one group prestest postest design

Back 142

--changes in pretest to potest may b doue to history, maturation, instrumentaiton, testing, stat regression
-when looking at changes the treatement caused so confouned to mulitple explanations

Front 143

one shot case study

Back 143

-one group is given a treattment followed by a test

Front 144

static group comparison design

Back 144

-tow group s but p r not assigned to the groups at random the dashed lines between the two gorup indicate they r intact groups
-in this there is a threat to internal validity called selecitons
-with pretest no way of knowing if the two groups r similar

Front 145

quasi experimental desings

Back 145

-r of intermediate value for eploring casue and effect should b used wen it is not possible to assign p at randiom
-nonequivalent contorl group design (has two intact groups)
-equivalent time sample design

Front 146

nonequivalent contorl group design (has two intact groups)

Back 146

subj to threats of marality, selection, interaction of sleciton , history
-superior to the preexperimental design
-often use some form of matching to increase the validity of results

Front 147

equivalent time sample design

Back 147

-has only one group or possible only one participant
-treatment conditions r alternated
-p are both the control and the experimental
-strong likelihood of muitple treatment interference

Front 148

ABAB

Back 148

- in psych a signle group design in whcih the treatments r alternated
-A and B indicate alt terematnets
-multiple observations before treatment is given called the baseline

Front 149

confound

Back 149

-is a source of confusion regardingthe ecplanation for a given diff
-hawthorne effect is an imporantce of confounding

Front 150

hawthorne effect which can b thought of as the alt effect

Back 150

-found in psych studies at hawthorne plant of western electric
-workers productivity went up wen the lighting was increased
-wen lighting decreased productivity also went up
-repsonding to the attention from the researchers
-two explanations for the diff observed

Front 151

to control hawthorne effect some researchers use three groups

Back 151

- an experimental group
-control group
-a control group that recieves no special attention

Front 152

the john henry effect

Back 152

-the possiblit that control groups might become aware of the infereior status and repsond by truing to outperform the epxerimal gorup thus should conceal

Front 153

blind procedure

Back 153

-when they do not diclose to the p whether they r recieving an active or inactive substance

Front 154

double blind experiment

Back 154

-neither the p nor the individual dispending hte drug kno wthich is the active drug

Front 155

demand characterisitiucs

Back 155

-b a source of confounding
-ais a cure that lets p kno that the expevted outcome of an experiment

Front 156

descriptive stats

Back 156

-summarize the data so they can easliy b comprehended

Front 157

frequency distribution

Back 157

-he or she can easily see how the scores r attributed

Front 158

inferential stats

Back 158

-purpose is to help researchers draw inferences about the effects of samping errors on athe results that r desvcribe dwith desciptive stats
-help researchers to make generalizaitons about the charactersistic sof pop on teh basis of data obtatined by studying samles

Front 159

margin of error

Back 159

-an inferenctial sta twidely reported in the mass media

Front 160

sig tests

Back 160

-important family of inferential stats conssits of this
-help resaerchers decide whether the diff in descrpitive stats they id r reliable

Front 161

if researchers dont sample

Back 161

dont need inferential stats

Front 162

parameters and stats

Back 162

a way to distinguish between values obtaine from a siample of values obtained from a census
-use parameters for values from teh census
-use stats for values from the studies in which samples were examined

Front 163

samples yield

Back 163

stats

Front 164

pops yield

Back 164

paramteres

Front 165

null hypothesis

Back 165

-version a: nthe observed diff was created by sampling error
-version b: there is no true diff between teh two groups
-verison c: the true diff between teh two groups is 0

Front 166

sig tests

Back 166

-determine the prob that the null hyp is true
-p stands for probablity
-if .05 or less then the null hyp is rejected
-if not rejected remains a possible explanation does not mean accepting as only explanation

Front 167

stat sig

Back 167

-another way to state that rejected the nullhyp

Front 168

four scales or levels that researchers measure

Back 168

-lowest nominal(naming level)-ex name marital status
-ordinal
-interval
-ratio

Front 169

ordinal

Back 169

-place p in order from high to low

Front 170

ratio and interval

Back 170

measre the amoung of diff among p
-equal distance among the scores they yeild

Front 171

ratio scale

Back 171

-has an absoute zero ex weight

Front 172

interval

Back 172

-no absolute zero so scores obtained from obj measures such as multiple choice test

Front 173

f

Back 173

frequencies

Front 174

N

Back 174

number of cases
-can b converted to percentages by dividding the vumber who prefer each candditte theot eh nubmber of the pop and multiply by 100
-wen reporting percentages want to report the underlying numbers

Front 175

univariable analysis

Back 175

-analyzing how p vary so se only one varialbe

Front 176

bivariate anaylsis

Back 176

-examine the relaitonshi pbetween two nominal variables

Front 177

porportions

Back 177

-sometimes reported instead of percentages in repors
-harder to comprehend then percentages
-good idea to convert mentally to percnetages

Front 178

null hypothesis

Back 178

asserts that any observed diff was created by random sampling errors

Front 179

chi square

Back 179

-ususal test of null hypothesis for diff between frequencies ie the number of cases or n
-symbol is x squared

Front 180

df

Back 180

degree of freedom

Front 181

value of chi square and degree of freedom were calculated

Back 181

-soley to obtain the probablity that he null hyp is correct
-chi square and degree of freedom r not descrptive stats focus on them as steps to find p

Front 182

bivariate

Back 182

-analysis that classifies p in terms of two variables in order to examine th erelatinship between the two variables

Front 183

type II

Back 183

-name of error researchers make wen they fail to reject the null hyp wen infact it is an incorrect hyp

Front 184

type I

Back 184

-name fo error researchers make wen they reject the nul hyp wen infact it is a correct hyp

Front 185

normal curve

Back 185

-bell hsaped curve

Front 186

skewed

Back 186

-if skewed to the right meaning tail is to the right has a positive skew

Front 187

the mean

Back 187

the most frequently used average
-is the balance pt in a distribution (positive and neg deviations balance eachother out
-to compute add up the scores then divide the numbers of the scores
-symbol for it is M(mean of pop), m (mean of sample)
-x with a line on top pronouned x bar
-the value around which the deviation sum is zero
-when mean is subtracted from all other scores the deviations r obtatined
-if deviations dont sum to zero the statisti is not eh mean

Front 188

major drawback of the mean

Back 188

-it is drawn in the direction of extreme scores

Front 189

median

Back 189

-avverage used wen distribution is highly skewed
-middle score
-to get one put scores in order from low to high

Front 190

mode

Back 190

-third average
-most frequently occuring score
-sometimes used in formal reporting but is very selfom reported in foraml reports of reserach
-not always a unique value

Front 191

synonym for average

Back 191

-measure of central tendecy
-used in resaerch and stat txtbooks wen averages discussed

Front 192

often a set of scores is described with only two stats

Back 192

- the mena to describe its average and the standard deviation (symbol S or SD for a pop and s or sd for a sample)
-to describe variablity

Front 193

variability refers to

Back 193

-amount by which p vary or differ from eachotehr
-smaller the variability smaller the standard deviation

Front 194

standard deviation

Back 194

if a distribution is normal 68 percent of the p inthe distribution lie within one standard deviation unit of the mean

Front 195

median

Back 195

- has 50 percent of the score above it and 50 percent of the score below it
-used to describe the average wen the data is ordianl or the distribution is highly skewed
-with this it is customary to report he range or interquartile range as a measure of variability

Front 196

range

Back 196

highest score minus the lowest score

Front 197

measurement theory indicates that

Back 197

-the more extreme the scores the more unreliable it is or the more errors
-to get aroudn this prob use a modified version of range -inerquartile range =the range between two quaters

Front 198

to calculat ethe interquarticle range

Back 198

-first divide the disctiburtion into two quaters

Front 199

pearson product mometn correlation coefficient

Back 199

-most widely used coefficient(wen want to examine the realtionship betwen two quantitative sets of scores)
-symbol is r ussually called pearson r - when relationship perfect =1.00

Front 200

direct relationship

Back 200

- alsco called positive relationship

Front 201

inverse relationship

Back 201

-negative relationship
-those who are high on one variable are low on the other

Front 202

coefficient of determination

Back 202

- pearson r has to be turned into this to rep correlation in terms of percentages
-symbol is r squared ( to compute simply square r)

Front 203

t test

Back 203

-often used to test teh null hyp regarding the observed diff between two mean s

Front 204

what causes a t test to yield lwo prob

Back 204

-sample size
-the size of the diff between means
-teh amount of variation in teh pop

Front 205

sample size

Back 205

-the large rthe sample the less liekely that an observed diff is due to teh sampling error
-other things being equal the larger sample teh lower the value of p

Front 206

the size of the diff between means

Back 206

-the larger the diff the less liely tthat teh diff is due to sampling errors
-so other things beign ewual, the larfer the diff between teh two means, the lower the value of p

Front 207

teh amount of variation in teh pop

Back 207

-wen a pop is very homogeneous there is less potetntial for sampiling error
-other things equal , the smaller the variability , the lower the value of p

Front 208

anova

Back 208

-analysis of variance
-yields a stat called F as well as degrees of freedom , sum of squares, mean square, and a p value which indicates teh prob that the null hyp is correct
-doesnt indicate whichof the three r sig

Front 209

t test and anova

Back 209

-base don same theory and assumptions
-wen two means r compared both tests yield the same valur for p
-1 t test =comparte two means
-anova-compare a number of means

Front 210

mutliple comarsion test

Back 210

indicates which of the diffs is sig
-

Front 211

one way anova

Back 211

-also called single factor anoca
-wen teh p have been classified in only one way

Front 212

main effect

Back 212

-in a two way analysis of variance

Front 213

even a a small dif can b

Back 213

a stat sig diff
-true becasue stat sig deteremine only the likelihoo dthat a diff is due to random error not wheter ist is a large diff

Front 214

practical sig

Back 214

-- frist sp in determiening the stat sig of a diff is to determine this
-involves 5 considerations

Front 215

involves 5 considerations

Back 215

--cost in relation to the benefit or cost benefit analysis
-crucial diff-one that results in a crucial increase or decrease
-client acceptablility-
-public and political acceptabiity-
-ethical and legal implications

Front 216

to select a purposive sampling

Back 216

-used in qualititative research
-must first id a research topic of interest and then seek individuals who are likel yto have releveant info

Front 217

qualitative researchers

Back 217

- make subj judgmeents regarding the individuals to select base don the liklihood that they will b able to provide the needed info

Front 218

purposive sampling is the correct term to use wen

Back 218

-teh researchers seek p who fit in to a braod category

Front 219

criterion purposvie samping

Back 219

-when there r a number of criteria to b applied in a slection of a sample the sampling tech is called this

Front 220

demo info

Back 220

should b collected so help consumber of research to see the p

Front 221

sometimes qualitative research r silent on the issue of how

Back 221

there samples were selected and provide lil demo info i nther ereprots

Front 222

somtimes qualitative researchers r interested in

Back 222

-researching a prob that exists in a particular locaiton or instituation

Front 223

wen qualitative researchers r interested in rsults that may havve broader app

Back 223

-they should seek p froma variety of sources

Front 224

one way to obtain data from divese sources is to

Back 224

-renta list of names and addresses of potential p

Front 225

qualitative usually use fewer

Back 225

p than quantitaive
-qualitative data collection is more expensive and tiem consuming
-

Front 226

semistructured interview s

Back 226

-the most widely used tpe fo mesasur efor collecting data for qualitative research
-usually face to face interciews usually recorded
-shoudl b an interview protocol

Front 227

interview protocal

Back 227

-consisting of written directionf for conducting the interview as well as a standard set of predetermined question tob e asked of all the p
-should b pilot tested with at lest a few individuals who will not b p
-also reviewd by epxerst in teh aresa

Front 228

intial question should b to

Back 228

-establish rapport

Front 229

semistructred refers to

Back 229

- dont have to ask all the predetermeind quesitons
-can b reworded
-if to terse can add
-can prob with add questions

Front 230

interciewers

Back 230

- must b practiced
-unbiased can achieve this through self disclosrue

Front 231

iissue of reality

Back 231

-raised wen the use of interview datat to explore complex issues

Front 232

phenomenological approach

Back 232

-examining perceptions is this way to acquiring knowledge

Front 233

focus groups

Back 233

-sometimes used by qualitative research
6-12ppl
-lead by faciliatator
-lasts abouta n hr
-2 or more focus groups in a project
-reveals the evolution of perceptions in a social context

Front 234

in qualitative research an important distinction is made between

Back 234

-nonparticipant observation and participant observation

Front 235

field research

Back 235

researchers who use obersvational methods often refer to their research as this

Front 236

ethnography

Back 236

-wen the focus is on a cultural issue

Front 237

data triangulation

Back 237

-in qualitative research used to id the dependablity and trustworthines of their data
-uses multiple sources for obtainign data

Front 238

methods triangulation

Back 238

-in qualitative researrch used to id the dependablity and trustworthines of their data
-two metods were employed to collect data from one group of p

Front 239

researcher triangulation

Back 239

-which reduces the possbiity that the results of qualitative research reps only the idiosyncratic vies of one individual -is basically resarech team

Front 240

research team

Back 240

-important tech to assure the quality of qualitative research

Front 241

team of researchers with diverse background

Back 241

--can sometimtes b helpful

Front 242

checkiing the accfuracy of a transcription

Back 242

-ensures the quality of th edata
-can do this by having an independent person compare teh adiorecording with teh transcripts or have the person chekc ony a sample of the transcript

Front 243

interoberver agreement

Back 243

-wen researchers disagree discuss diff unti they reach a consensus

Front 244

peer review

Back 244

-examine the process used to collect teh data and the rsulting data in teh conclusions and provide feedback -sometimes called the auditor

Front 245

member checking

Back 245

- can also enhance dependablity
based ont eh idea that the p r memeber of the resarch team
-by having the p/members revies the results of the analyiss researchers fcan determine wheterh their results ring true to the p

Front 246

grounded theory

Back 246

-theory refers to inductive mehtod of analyiss that can lead to theories of behavior

Front 247

grounded theory approach stars with

Back 247

open coding
axial coding

Front 248

open coding

Back 248

-the transcripts of the intervies r examined for thedisticnt and seprate segments which r id by the type and coded with th individuals name

Front 249

axial codin g

Back 249

-the transcipts of theinterviews and any other data source such as memeos written during data colleciton r reexamined with the purpose of id relationships between the categroes and themes id during the open coding
-impranta relationships

Front 250

important relationships

Back 250

1. temporal (x usually proceeds y in time)
2. causal (x caused participants to do y)
3. associational ( x or y ussualy or alwyas occur at aobut he same time but r not beleived to be causally connected
4, valence participants have stronger emotioanl connections to x than to y
5. spatial x and y occur in teh sam e place or x and y occur in diff places

Front 251

core category

Back 251

-the main overarching category
-attempt to describ eth eprocess that lead t ot eh relatinship id in the previoeu stage of the anlayis s

Front 252

constant comparison

Back 252

- key element thorught the analysis of data using the grounded theory appraoch
-constantly comparing each new element of the data with all previous elemts that ahve been coded in order to establish and refine categoreis

Front 253

consensual qaulitative research

Back 253

-CQR
-qualitative research
-are inductive
-wants researchers to arrive at a consensus of the meaning of the data collected

Front 254

step of CQR is to

Back 254

code into domains
-develop core ideas within domains
-cross analysis

Front 255

code into domains

Back 255

-segmenting the data into groups according to the topic they cover
-after all member of research team done then discuss and refine domains until they reach unanimous agreement on tehm

Front 256

develop core ideas within domains

Back 256

-writing short summaries that reduce the original ides of p to fewer words

Front 257

cross analysis

Back 257

-the core ideas r grouped into categories based on similiarities
-higher levels of genralization because of this

Front 258

in CQR there may be

Back 258

-an external stability check done by examining data in add to the interview transcripts

Front 259

internal stability

Back 259

-is ecamined in CQR by determining the extent to which each category was general, typical, or varient

Front 260

genral

Back 260

domains that apply to all p

Front 261

typical

Back 261

those that apply to half or more

Front 262

varienat

Back 262

apply to less then half but more than 2

Front 263

auditor

Back 263

-CQU requres one
expert that review sth work of the research team after each major stepp

Front 264

ina report on a CQR study the following should b briefly described

Back 264

-the credentials of the auditor, steps at chich the auditor supplied feeback, wheter the changes were made on the basis fof the audiotrs feedback