Brainscape's Knowledge GenomeTM

Browse over 1 million classes created by top students, professors, publishers, and experts.


See full index

Flash Card Library > Research Methods > Flashcards

Research Methods Flashcards

1
Q

Explain reliability and validity?

A

Reliability– is the consistency of the measurement (the umpire being consistent even if it is not accurate)
“Reliability on the scores! Not the test! TESTS DON’T HAVE RELIABILITY”

Validity- is the accuracy of the measurement or instrument (when the umpire calls a strike and its a strike the call is valid)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What are some tests to measure reliability?

A

(1) Test re-test
(2) Alternate forms
(3) Internal consistency
(4) Interrater reliability

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Define and/or explain the four tests to measure reliability?

A

(1) Test re-test: repeat measurement and should receive similar results
(2) Alternate forms- prepare alternate versions of the test covering the same content but with different items
(3) Internal consistency reliability: split half reliability; two halves of the test meet the standards of parallel forms
(4) Interrater reliability: obtain ratings from multiple sources, i.e. two researchers

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Give an example of observed score vs. true score; and what is the formula that depicts this phenomena?

A 
My actual height is True Score,
The height measured on the wall or another person is Observed Score
X- observed score
t- true score
e- error
X= t + e
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Name and explain the measures of central tendency?

A

Mean- the average of scores, Sum of X over n.
Median- middle score
Mode- the most frequent
The mean takes into account the magnitude of all the scores, in contrast the median takes into account only the numbers of scores and the values of the middle scores.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

In a normal curve how are the percentages distributed?

A

Mean is at the center, 34% on either side- represents one sd from the mean, then second line is 14%, then finally 2% at the tails.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Where do the mean, median, and mode fall on a positively skewed distribution vs. negatively skewed distribution?

A

Positively skewed: starting from far right to left: Mean, Median, Mode
Negatively skewed: starting from far left to right: Mean, Median, Mode

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What are the three measures of variability?

A

Range, Variance, and Standard Deviation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What is the range?

A

R= H - L, highest minus lowest scores. “range of scores are from 1 to 5”

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What is variance?

A

σ2 = Σ(x1 –m)2
N - 1

The average deviation of all the numbers from the mean in squared units.
Whenever we are doing any research we want to explain the variance in people.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What is standard deviation

A

SD = square root of Σ(x1 –m)2

N - 1

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What are the five steps to figure SD?

A

Step1: find mean: Σ X

Step 2: subtract the mean from each number:
Σ (x – x bar)

Step 3: square of each number (to get rid of negative numbers): Σ (x – x bar)2

Step 4: Variance: Σ(x1 –m)2
N - 1

Step 5: SD: square root of Σ(x1 –m)2
N - 1

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What is a Z score?

A

Raw score that has been transformed into standard deviation units;
A z-score tell you how many standard deviations a raw score is from the mean
i.e. when looking at GRE scores (a z-score of +2 means a score 2 standard deviation above the mean.

z= x - xbar/ sd

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

What is statistical significance?

A

Researchers claim their finding is statistically significant when they do not believe that their observed result was due only to chance or sampling error.

Answers the questions: is there a relationship between the IV and the DV? Is it greater than 0?

Rejecting the null hypothesis

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What is effect size?

A

A measure of the strength or magnitude of a relationship between the IV and the DV

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

What is regression analysis?

A

A set of statistical procedures used to explain or predict the relationship between a dependent variable and one or more independent variable.

Regression analysis is also used to understand which among the independent variables are related to the dependent variable, and to explore the forms of these relationships.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

What is the regression equation formula? Both with and without standardized scores?

A

Non-standardized scores: y hat = a + bx
With standardized scores: y hat = b1x1 + b2 x2
y hat is the predicted value of the dependent variable
a is the y intercept
b is the regression coefficient or SLOPE; the predicted change in Y given a 1 unit change in x; the amount we will see change in salary from 2.5 gpa to 3.0 gpa.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Describe the components of the regression summary table?

A

SE Variable R R2 delta R2 Beta b
Accult Stress 0.4 0.16 0.1
Fam Support 0.5 0.25 0.09 0.1
.4 represents that correlation coefficient between SE and AS?
.5 represents the correlation of the combined AS and FS?
What does delta R2 represent? The change between R and R squared (.25 - .16)
Beta column represents the standardized score? And b represents the unstandardized?
Beta’s indicate the unique correlation between each variable and the criterion

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

When you are doing regression analysis, you have to decide how to enter the x variables. What are three entry methods/procedures?

A

1) Simultaneously- use this when you want to find how well the variables collectively predict y and which is more important
2) Step-wise- usually the default; use this when you want to show the best set of the predictors
3) Hierarchical- use this when you want to find out how the other variables relate, i.e. I know Acculturative Stress relates, and I want to find out if family support adds anything

20
Q

What is T-test?

A

Statistical test used to determine whether the difference between the means of two groups is statistically significant.

21
Q

Selecting people for your study is an important consideration. What is range restriction?

A

the more restricted or narrow your scores are on your measures vs your population, the lower reliability. (i.e. African American Women and two clinic where African American Women are recruited- they already have been identified as at risk or receiving treatment so the range will be narrow)

22
Q

Describe Type I and Type II error

A

Null hypothesis is TRUE or FALSE (should be rejected)
FAIL to reject it or REJECT it…
TYPE I: When the null hypothesis is TRUE, but it is REJECTED– False Positive, claiming positive outcome or finding when it really was not!
TYPE II: When the null hypothesis is FALSE, but it is NOT rejected. False Negative, claiming there wasn’t any findings when there was.

23
Q

What is the error rate alpha:α?

A

α error rate was decided by convention that if the probability of making a Type I error is 5 chances out of 100, .05, or less that’s good, but 6 chances out of 100 is not!

24
Q

What does t distribution tables tell us?
lower case v?
2Q?

A

lower case v= degrees of freedom; degrees of freedom for T-test is n - 2
2Q= two tailed test
This table tells us if your T-test yields the corresponding number or higher you can reject the null hypothesis

25
Q

Explain the experimental wide error rate?

A

You can have multiple tests, each set at alpha error rate of .05; the experimental wide error rate asks what is the probability that at least one of the significant results is a Type I error.

26
Q

What are three key ingredients to a good/quality experiment?

A
  1. Ask clear questions
  2. Collect data to address your question
  3. Analyze data for the question…And Stop!
27
Q

Define power and how do you increase your power?

A

Power is the likelihood of rejecting the null hypothesis when it is false. (Avoiding Type II error)

  1. Select measures that have SCORES that look like they are highly reliable from people SIMILAR to your study
  2. Get people in your sample that increase the RANGE of scores.
28
Q

Name threats to INTERNAL VALIDITY? Remember if comparing groups, this is very critical.

A

In general running a “gold standard” experiment (Random assignment, pre-test post test control group) will eliminate most of these:

  1. Maturation effects- time elapsed
  2. History effects- environment changes
  3. Attrition
  4. Testing
  5. Regression to the mean- when taking two tests, the second test has the likelihood of regressing to the mean
  6. Selection
29
Q

What is the Gold Standard experimental research design?

A

Random assignment- Pretest posttest control group design

30
Q

What is a quasi-experimental design?

A

Defined by non-random assignment

31
Q

Name three quasi-experimental designs and describe them?

A
  1. Non-equivalent comparison-group design
    01 X1 02
    - - - - - - - -
    01 X2 02
    Analyze by (a) comparing the pre-test to post-test differences scores of the two groups or
    (b) comparing the experimental and control group’s post-test scores after they have been adjusted for any differences that might exist on their pre-test scores using ANCOVA.
  2. Interrupted time-series design (TIME Series design)
    Mult. pre-test….. treatment….. multiple posttests
    01, 02, 03, 04 X1 05, 06, 07, 08
  3. Cohort design
32
Q

Lessons from Wilkinson et al (1999): Variables

A

explicitly define the variables in the study

33
Q

Lessons from Wilkinson et al (1999): Instruments

A

Always summarize the psychometric properties of the scores on the measures used in research: reliability, validity, and other…Test is neither reliable nor valid!

34
Q

Lessons from Wilkinson et al (1999): Choosing minimally sufficient analysis

A

If the assumptions and strengths of a simpler method are reasonable for your data and research problem, USE IT!

35
Q

Lessons from Wilkinson et al (1999): Power and sample size

A

Provide information on sample size and the process that led to sample size decision. It is important to show that the effect size estimates used to calculate power have been derived from previous research and theory…

36
Q

Lessons from Wilkinson et al (1999): Multiplicities

A

We often encounter many variables and many relationships. Remember: if a specific contrast interests you, examine it. If all interest you, ask yourself why?

37
Q

Lessons from Wilkinson et al (1999): Effect Sizes

A

ALWAYS provide an effect size estimate when reporting a p value. Always present an effect size for primary outcomes.

38
Q

Lessons from Wilkinson et al (1999): Internal Estimates

A

Internal estimates should be given for any effect size involving principal outcomes. Comparing confidence intervals from a current study to intervals from previous, related studies helps focus attention on stability across studies.

39
Q

Lessons from Wilkinson et al (1999): Random assignments

A

For research involving causal inferences, the assignment of units (e.g. subjects) to levels of causal variables (e.g. treatment and control groups) is critical.

40
Q

Lessons from Wilkinson et al (1999): Non-random assignment

A

Need to minimize the effects of variables that affect the observed relationship between a causal variable (e.g. treatment) and an outcome. Such variables are commonly called confounds or covariates.

41
Q

Lessons from Wilkinson et al (1999): Causality

A

Inferring causality from nonrandomized designs is a risky enterprise. Researchers using nonrandomized designs have an extra obligation to explain the logic behind covariates included in the design and alert the reader to plausible rival hypotheses that might explain their results.

42
Q

Lessons from Wilkinson et al (1999): Interpretations

A

Novice researchers err either by overgeneralizing their results or, equally unfortunately, by over-particularizing. Be sure to explicitly compare the effect sizes in your study with the effect sizes reported in (or calculable from) related previous studies.

43
Q

Lessons from Wilkinson et al (1999): Conclusions

A

Note the shortcomings of your study. Remember, however, that acknowledging limitations is for the purpose of qualifying results and avoiding pitfalls in future research. Recommendations for future research should be thoughtful and grounded in present and previous findings.

44
Q

Convert T-test to r (formula)?

A

r= Square root of t squared /t squared + df

45
Q

Convert T-test to d (formula)?

A

d= 2 t /Square root of df

46
Q

Convert d to r (formula)?

A

r= d/ Square root of d square + 4

47
Q

Convert r to d (formula)?

A

d= 2r/ Square root of 1 - r squared

Flash Card Library (15 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2024 Bold Learning Solutions. Terms and Conditions