Educational Research

Front 1

A scietific theory must contain three componants

Back 1

1. Explain
2. Predict
3. Stimulate new research ideas
4. Falsafiable
5. Replicable

Front 2

A Construct must:

Back 2

1.Have to match your theory
2.not testable/untappable
3.Biggest part either proves or disproves your study.
4.Cannot test abstract theories, constructs are abstract, cant be observed directly
5.Used to summarize observations and provide explanations for behaviors
6.Must be given in operational definitions

Front 3

An example of a construct is:

Back 3

IQ, motivation, drive, ambition

Front 4

Operational definition:

Back 4

1.Ascribes meaning to the construct by being more specific and telling them what you are going to do
2.Bridges gap between observable and theoretical, you have to be able to observe it happening
3.Specific and delineates a term so we are all on the page
4.Allows you to measure abstract concepts
5.Choice of the researcher

Front 5

Variable from the construct:

Back 5

1.are what you test
2.characteristic that takes on a different value or scores

Front 6

Types of Variables:

Back 6

1.Categorical
2. Discrete

Front 7

Catagorical

Back 7

classified into groups, ethnicity, language, occupations, and schools you attend (all FIU students) ect…

Front 8

Dichotomous variable

Back 8

simplest pass-fail, male-female test scores, height, age

Front 9

Independent

Back 9

-discrete
(tested by t-test, ANOVA, MANOVA) can only take on a finite number of values ,usually all qualitative variables

Front 10

Dependant

Back 10

likely to be continuous variables that can take on an infinite number of possible values

Front 11

A Hypothesis must contain: (5)

Back 11

1.Integrates information-tentative statement about how variables are realted
2.Stimulates research
3.Provides testable statement
4.Provides direction
5.Provides framework to report findings

Front 12

Name some limits in Ed Research

Back 12

1.Human cohort effects
2.Observation
3.Replication
4.Control
5.Interaction effects b/w observers and pts
6.Measurement errors

Front 13

Population

Back 13

Can’t test it to large-abstraction
not researchable

Front 14

Sample

Back 14

1. should be representative of the population you are trying to generalize towards

Researchable

Front 15

Reasoning: What is the difference between inductive and deductive reasoning?

Back 15

1.Deductive: general-specific (theory to constructs)

2.Inductive: specific-general (specific event to theory)

Front 16

The Scientific Approach uses

Back 16

inductive-deductive reasoning
1.You start small then work your way up…..specific to generalizable theory

Front 17

Name the componants of scientific research

Back 17

Components
Problem/Question
Observe/Gather information (Literature review)
Form hypotheses
Establish method for collecting data
to test hypotheses
Analyze data
Interpret data/Draw conclusions

Front 18

Limitations of Education Research

Back 18

Human subjects = human error & interaction effects between observer & subjects
Complex constructs; difficult to operationalize
Difficult to…
…conduct observations …replicate …control
measure variables (tools are not good)

Front 19

Types of Research
Empirical Research consists of...

Back 19

1. requires a norming group of students and their average mean-then becomes the criterion mean to which all others are compared to. Sampling of the population of interests average scores.

Front 20

Rationalization/reasoning:

Back 20

This is the old Descarte-French Philosophy theory of "I think therfore I am:
It is ********, Socrates and Aristotle philosophy does not account for real events, acceptable in the context of the historical times it was dervived from irrelevant now.

Front 21

Basic Research

Back 21

aims to obtain empirical data used to formulate, expand, or evaluate a thoery, not solve a real world problem

Front 22

Applied

Back 22

aims to solve an immediate problem-practical

Front 23

Method or statistical analysis is__________a personal choice.

Back 23

Method is NOT a personal choice…the problem and questions guide the research

Front 24

Quantitative
Pupose:
approach:
design:
tools:
samples:
analysis:

Back 24

P-generalize, predict behavior, provide causal explanations
A-begin with h1, thoery grounds study, manipulation of variables, deductive then inductive, analyze discrete componants, looks for norm,reduced data to #'s, written in prcises abstract lang.
D-focus quantity(how much,many) experamental, empiracal, stats focused, predetermined,
s-random best
t-inferential stats, measurment scales, scores, questionaires, checklists, exams

Front 25

Qualitative
Pupose:
approach:
design:
tools:
samples:
analysis

Back 25

p-contextualize findings, interpret behavior, understand perspectives
d-ends with h1, (emergant data)natural context, inductive-deductive, larger patterns, looks for complexity, relies on words-min use of numerical data descriptive holistic lang.fieldwork, ethnographic
t-observations, case studies, interviews
s-perposive
a-quality (nature or essence)

Front 26

Mixed methods: work but are hard to do-why?

Back 26

b/c can't use a qualitative study and at the end add data analysis and call it quanitative, post hoc

Front 27

Ethnographic designs: are q___, and study__________.

Back 27

quantitative and study cultural aspects or society as a cultural group to gain a holistic picture of a cultural group in-depth interviews, prolonged pt observations needded

Front 28

What is the difference between a inductive and deductive H1?

Back 28

Inductive-specific to general
Deductive derived from theory

Front 29

A H1 is based on the relations b/w_____, and on the operational constructs devired from a ________.

Back 29

variables, theory

Front 30

Know how a H1 is constructed the order and meaning behind it......inductive and deductive

Back 30

Drawn out

Front 31

Theories should do 5 things

Back 31

relation between constructs
Organize, explain, predict, stimulate,Parsimony

Front 32

Test the Null hypothesis…why?

Back 32

Innocent until innocence is doubted beyond reasonable moral judgment
Easier to find support for a reason (Ha) when demonstrating the opposite (Ho) to not be true

Front 33

Type I; α error;

Back 33

false-positive = reject Ho when Ho is actually true (find man guilty when innocent)

Front 34

Type II; β error;

Back 34

false-negative = fail to reject Ho when Ho is actually not true (find man innocent when not innocent)

Front 35

population parameters are used when writing a H1 b/c

Back 35

defines the populations triats or characteristics

Front 36

Reject vs. Fail to reject Ho

Back 36

have to know when to use them based on inferiential stats

Front 37

Support vs. Fail to support Ha

Back 37

have to know when to use them based on inferiential stats

Front 38

Likert scale

Back 38

5-point strongly agree-disagree doesn't allow for individual answers-forced response

Front 39

Reverse scales

Back 39

not face valid-score it opposite to get data needed

Front 40

Coding systems
Sign coding
Time coding

Back 40

SC-this is frequency of the observed behaviors in a certain time frame
TC-did the behavior occur within a ten minute time period marked by little checks

Front 41

Errors/Bias----
Halo-
Observer bias-
Observer Effects-

Back 41

H-raters genrl impressions influence rating given to subjects behaviors
OB-observer personal attitude/values affect observation and interpretation of data
OE- impact of observer has on pts. performance....see social desirability effects

Front 42

Reliability

Back 42

Reliability – repeatability, consistency

Front 43

Validity

Back 43

– measure what it says it does (instrument and “use” of instrument/data)

Front 44

Threats to validity

Back 44

Construct under-representation
Construct irrelevant variance

Front 45

Validation

Back 45

Content/Face

Front 46

Criterion validity

Back 46

Concurrent, Predictive

Front 47

Construct validity

Back 47

Convergent, Discriminant

Front 48

Descriptive Statistics
Scales of measurement

Back 48

Nominal-catagorical

Ordinal-ranking orders

interval-intervals between measures, or numerical data inbetween points, no true
zero point

Ratio-interval and ratio data, true zero point

Front 49

Distribution Frequency

Back 49

how many time it appears in the data (f)

Front 50

Histogram (Polygon)

Back 50

Histogram (Polygon)-bar graph and points are the .5 ranges inbetweenCurve shapes

Front 51

Symmetrical

Back 51

how is the data presented in a bell curve, this is symmetrical all data is evenly distributed across the curve

Front 52

Measures of Central Tendency:

Back 52

Measures of Central Tendency: Mean-average, Median-the mid point, Mode-frequency

Front 53

Variability Range R

Back 53

R=Xh-X1+I 1-2-3-4-5-6 = diff b/w upper and lower limits mean 6-2+1=5

Front 54

Restriction of Range-deviation scores

Back 54

d=x-m so m=5 x=6 then the deviation is 1

Front 55

Variance:

Back 55

Variance: how much do your scores vary in the bell curve or vary from the mean of the average scores, lots of types of variation

Front 56

Standard Deviation

Back 56

standard deviation is how many sd’s does your score deviate from the average mean, meaning does it deviate 1sd or 2sd from the mean above or below

Front 57

For research purposes, “standardized” indicates

Back 57

that the test measures what it says it measures, does so consistently, and have comparable norms. Therefore in research, “standardized test” is one that has been standardized against a representative sample. (In common speak, “standardized test” has come to refer to any test that is administered and scored in a consistent manner.)

Front 58

Standard scores/percentile – relative position to others

Back 58

E.g., one who is at the 90%ile scored higher than 90% of the “norm” sample; or a test score which is greater than 90% of the scores of people taking the test is said to be at the 90th percentile

Front 59

Normed/Norm referenced

Back 59

– compared “normal” or typical population (gathered by a sampling of the population)

Front 60

Non-standardizedCriterion referenced – what an individual can do

Back 60

– what an individual can do

Front 61

Some tests provide both

Back 61

criterion and normed scores

Front 62

Range The overall numerical distance of a test is

Back 62

1.0 – 100 is the range of a typical exam.
Restriction of range – range is not wide enough to allow for participants to vary freely
Variability – spread or dispersion of a measure

Front 63

Ceiling effect

Back 63

Ceiling effect – range is too low on upper end
Does not allow for variability at the higher end.

Front 64

Floor effect

Back 64

Floor effect – range is too high on lower end Does not allow for variability at the lower end.

Front 65

Likert scale
Reverse scoring

Back 65

Likert scale – rated questions (SD = 1, D = 2, A = 3, SA = 4)
Reverse scoring – negative questions should be scored in reverse

Front 66

Observations: Watch behavior/s and record notes/data...how?

Back 66

Checklist – did behavior occur yes/no

Rating – rate behavior/s in terms of amount that occurred

Coding systems – edetermined behaviors to look for

Sign coding – mark when behavior occurs
Time coding – during set time, did behavior occur Combination

Front 67

the operational definition, which is what is used to determine the ______, must fits the ________.

Back 67

measure, conceptual definition

Front 68

Two problems that threaten validity

Back 68

Construct under-representation – assessment/measure is too narrow and fails to include important dimensions of the construct
Construct irrelevant variance – extent to which scores are affected by extraneous processes to the construct

Front 69

Validation:
1.Content

Back 69

– do questions relate to domain (judgmental validation); no numerical index; “face validity”

Front 70

2.Criterion

Back 70

a.Concurrent validity – new measure given with a given criterion measure at same time to determine if they relate; ability to distinguish between groups that it should theoretically be able to distinguish between.

i.A new math measure should be able to distinguish between those on the math team and those not on the math team.

b.predictive validity – new measure given now and correlates with criterion measure given at a later time point; ability to predict something it should theoretically be able to predict.

i.A new math measure should relate to (or predict) later grades in high school calculus.

c.(Validity coefficient (rxy) – is simply a correlation)

Front 71

3.Construct (2)

Back 71

does score on assessment measure underlying construct

a.Convergent validity is the degree to which concepts that should be related theoretically are interrelated in reality; the degree to which the operationalization is similar to (converges on) other operationalizations that it theoretically should be similar to.


b.Discriminant validity is the degree to which concepts that should not be related theoretically are, in fact, not interrelated in reality; the degree to which the operationalization is not similar to (diverges from) other operationalizations that it theoretically should be not be similar to..

Front 72

Construct Validity--

Back 72

How well one moves from construct to variable through operational definition.
How well one translates the theories/ideas into actual programs or measures.

Front 73

Two types:Construct Validity

Back 73

Convergent and discriminant: if you can demonstrate that you have evidence for both convergent and discriminant validity, then you've by definition demonstrated that you have evidence for construct validity. But, neither one alone is sufficient for establishing construct validity.

Front 74

measures of constructs that theoretically should be related to each other are, in fact, observed to be related to each other

Back 74

(that is, you should be able to show a correspondence or convergence between similar constructs)

Front 75

measures of constructs that theoretically should not be related to each other are, in fact, observed to not be related to each other

Back 75

(that is, you should be able to discriminate between dissimilar constructs)

Front 76

"high" do correlations need to be to provide evidence for convergence and how "low" do they need to be to provide evidence for discrimination?

Back 76

the convergent correlations should always be higher than the discriminant ones

Front 77

Descriptive Statistics

---discrete

Back 77

Discrete
Nominal – name only; no numeric meaning; gender, number on jersey
Ordinal – rank order; distance between rank is meaningless; first place, second place, etc.; 0=no h.s., 1 = ged 2 = h.s. diploma 3 = some college 4 = college degree

Front 78

Continuous -Interval

Back 78

Continuous Interval – can measure distance between numbers; but ratios are meaningless; the distance between 30 and 40 degrees is the same as between 70 and 80 degrees; however 80 degrees is not twice as hot as 40 degrees (though it is twice as large); in other words there is no true, meaningful, absolute 0 (in temp, 0 still has plenty of meaning).

Ratio – distance is meaningful, and ratios are meaningful; there is an absolute and meaningful 0; 80 inches IS twice as long as 40 inches.

Front 79

Distributions- Frequency

Back 79

X f fX cf
Histogram – continuous X (bar plot if discrete) (polygon connects dots instead)

Front 80

Measures of Central Tendency

Back 80

Tendency Mean, Median, Mode
Mean = arithmetic average
m= X1 + X2 + … + Xn
N

Front 81

Variability Range

Back 81

= difference between the upper real limit and the bottom real limit OR the number of intervals between the bottom number and the top number
R = Xh – Xl + I

Front 82

Deviation scores =

Back 82

scores that show difference between raw score and mean of distribution d = X – M