Brainscape's Knowledge GenomeTM

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


See full index

Stats 580 > Terminology > Flashcards

Terminology Flashcards

1
Q

Independent Variable

A

A variable thought to be the cause of some effect. In research, usually describes variable that researchers manipulated

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

Dependent Variable

A

Variable thought to be affected by changes to independent variable

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

Predictor Variable

A

A variable thought to predict an outcome. Basically an independent variable.

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

Outcome Variable

A

A variable thought to change as a function of change in a predictor variable. AKA dependent variable.

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

Hypothesis

A

A prediction about something

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

Between Subjects Design

A

An experimental design where different treatments used different organisms. i.e. Group A gets treatment A and group B gets treatment B

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

Within Subjects Design

A

An experimental design where different organisms receive more than one different treatment. i.e. Group A gets treatment A and treatment B

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

Categorical Variable

A

A variable made up of categories of objects. E.g. UK degrees 1, 2:1, 2:2, 3

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

Binary Variable

A

Type of categorical variable. Must fall into one of two distinct categories e.g. yes or no, pregnant or not pregnant

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

Nominal Variable

A

Where numbers represent names. E.G. sports jerseys. Number one not necessarily better than number two.

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

Ordinal Variable

A

Tells us things and the order that they occur in. e.g. bronze, silver, and gold medals

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

Continuous Variable

A

A variable that can be measured to ANY level of precision. e.g. time

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

Interval Variable

A

A variable consisting of interval data. Example a 5 star rating system where the interval between 2-3 and 4-5 are the same.

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

Ratio Variable

A

In interval variable, but ratios are also meaningful. e.g. a rating of 4 should be twice as good as a rating of 2, and 2 twice that of 1

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

Discrete Variable

A

A variable that only takes on certain numbers in a scale, usually whole numbers

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

Measurement Error

A

The discrepancy between the numbers used to represent a thing were measuring and the actual value of the thing were measuring

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

Validity

A

Did a test measure what it set out to measure

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

Reliability

A

Ability of a measure to produce consistent results when performed under different conditions

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

Correlational Research

A

A form of research where you observe what happens without interfering with it. Data will be analyzed to look at relationships between naturally occurring variables rather than statements about cause and effect

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

Experimental Research

A

Where one or more variable is manipulated to see their effect on the outcome variable

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Ecological Validity

A

Evidence that the results of a study, experiment, or test can be applied, and allow inferences, to a real world connection

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Systematic Variation

A

variation due to some genuine effect. Either the researcher doing something to all the participants in one sample, but not the other, or due to natural variation. Can be explained by model we fit to our data.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Unsystematic Variation

A

Variation that isn’t due to the effect we’re interested in. Variation that can’t be explained by whichever model we’ve fit to the data.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Counterbalancing

A

Process of systematically varying the order in which events are conducted. Example half the participants do condition A then B, other half does condition B then A.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Q

Research Question

A

A question that research sets out to answer

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
26
Q

What is normal distribution?

A

If a line was drawn through middle, would be the same on both sides. Bell shaped. Majority of scores lie around centre of distribution.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
27
Q

What is skew?

A

Lack of symmetry. Positive skew=scores at lower end, Negative skew=scores at higher end

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
28
Q

What is kurtosis?

A

Degree to which scores cluster in middle. Positive (leptokurtic) = pointy middle. Negative (platykurtic)=flat middle

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
29
Q

What is central tendency?

A

Where the centre of a frequency distribution lies

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
30
Q

Mean

A

Average score

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
31
Q

Median

A

Middle value. Or average of two middle most values

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
32
Q

Mode

A

Most frequently occurring number.

33
Q

Bimodal Distribution

A

Two numbers tied for most occurring number

34
Q

Multimodal Distribution

A

More than two number share highest rate of occurrance

35
Q

Statistical Model

A

A mathematical model that embodies a set of statistical assumptions

36
Q

What does “the mean is a statistical model” mean?

A

Used when the mean value is not a number that exists in the data set, but does represent the mean.

37
Q

What is variance?

A

An estimate of the average variability of a data set

38
Q

What is standard deviation?

A

The square root of variance. Small SD indicate values are close to the mean. Large=wider distribution

39
Q

What is the difference between populations and samples?

A

A population is an entire group of people with a given set of characteristics. A sample is a subset of a population.

40
Q

What is central limit theorom?

A

States when samples are large (>30), a small random sample will take the shape of a normal distribution regardless of where it is drawn from.

41
Q

What is a histogram?

A

A chart showing a frequency distribution

42
Q

What is a boxplot and what does it show?

A

Displays a five number representation of a data set. Minimum, first quartile, mean, third quartile, maximum.

43
Q

What is the standard error of the mean?

A

The standard deviation of a sampling distribution of a statistic. i.e. how accurately a sample represents its population.

44
Q

What is a 95% CI and what does it mean?

A

Represents a range (usually around the mean) in which 95 percent of the data falls in.

45
Q

What is bivariate correlation and when is it used?

A

A bivariate correlation is a correlation between two variables. It is used when one of the variables is dichotomous (can be divided into two distinct things. e.g. land mammals and aquatic mammals).

46
Q

What are part and partial correlations?

A

Partial: A correlation between two variables while others are controlled for their effects on both variables. Variation unique to those two only.
Part: Control for the effect of a third variable on only ONE of the two in the correlation, not both

47
Q

How do you compare independent and dependent correlations?

A

Independent: Convert coefficients to Zr (this makes the sampling distbn normal). Then calculate the Z-score of the the differences of the coefficients. Look up value of z to get one tailed distbn, multiply by 2 to get two tailed.

Dependent: Use a t-statistic

48
Q

What is linear regression?

A

Fits a linear model to minimize the residual sum of squares between the observed and predicted data set.

49
Q

What is the method of least squares?

A

Method of finding a line that best fits the data

50
Q

How do you assess the fit of a regression?

A

Compare the differences between the total sum of squares (SSt) and the residual sum of squares (SSR). This difference is the model sum of squares(SSM). If SSM is large, the regression model is different than the mean(SSt) (much betta). If SSM is small, not really any better than the mean (SSt)

51
Q

What are the model parameters of a linear regression?

A

Normal distrbn with a mean of 0 and constant variance

52
Q

What is covariance?

A

Measure of an “average” relationship between two variables.

53
Q

What is the central limit theorem and why is it important?

A

Says that when samples are large (30+), the sampling population will assume a normal distribution regardless of the shape of the population it was drawn from.

54
Q

What is the sampling distribution of the mean?

A

When repeated random samples of the same size are taken, the mean of all sample means is the mean of the population.

55
Q

Explain the logic of null hypothesis testing?

A

Assuming the null hypothesis is true, finding how likely the sample result would be if Hnot was true, and then making a decision

56
Q

What is a z-score?

A

Represents the difference between the data point and the mean

57
Q

How and when is a single sample t-test used?

A

Used to determine whether an unknown population mean is different from a specific value

58
Q

When is a dependent sample t-test used?

A

Used to determine if two conditions results differ for the same participant

59
Q

When is an independent sample t-test used?

A

Used to determine if two participants in different groups results differ

60
Q

What are the assumptions of an ANOVA?

A

1) Between group independence
2) within group sampling and independence
3) Normality
4) Homogeneity of variance

61
Q

What is an F score and how is it computed?

A

The F score is a measure of accuracy for a test. It is calculated using precision and recall.

62
Q

What is meant by between and within variance in the context of ANOVA?

A

Variability between the means is the variability between the means of each sample from the mean of the whole population.
Variability within the distributions is the spread of each sample.

63
Q

Explain the logic of contrast analysis.

A

Allows for pairwise comparisons of groups (like multiple t-tests), but corrects the level of significance to control the familywise error so that the error rate across all comparisons remains at 0.05.

64
Q

What are posthoc tests and how are they used?

A

They are a compilation of pairwise comparisons that compare all the different combinations of treatment groups.

65
Q

What are non parametric statistical tests and why are they used?

A

Non-parametric statistical tests are assumption free tests. Used when data breaks the parametric assumptions

66
Q

What is a main effect? Give an example.

A

A main effect is the effect of a variable in isolation. E.G. study examining attractiveness of dance partners after drinking varying amounts. Womens partners remain consistent, while mens decrease in attractiveness. This is main effect of gender.

67
Q

What is an interaction? Give an example.

A

An interaction represents the combined effect of two or more variables. E.G. with amount drank and partners attractiveness, differs within sex by number of drink, and between sexes.

68
Q

How do you post hoc a factorial ANOVA?

A

Analyse significant factors to see what is driving the interaction

69
Q

What is repeated measures ANOVA?

A

Used to assess a repeated measures test. (i.e. all participants partake in several measurements of experiment. ex. complete questionnaire after 1 shot, 2 shots, 4 shots, 6 shots.)

70
Q

How is the partitioning of variance different for a repeated measures ANOVA?

A

Within variance broken down into subjects (consistent with in subjects) and error.

71
Q

What is the assumption of sphericity and how can you test it?

A

The assumption that the relationship between experimental pairs is roughly similar (the level of dependence between experimental conditions is roughly equal). Assessed using Mauchly’s test.

72
Q

How do you post-hoc a repeated measures ANOVA?

A

Use a pairwise t.test with paired=TRUE

73
Q

What is mixed ANOVA?

A

An ANOVA assessing both between-group and repeated-measures comparisons. Need at lest 2 independent variables.

74
Q

What are the assumptions of mixed ANOVA?

A
  1. Dependent variable will be continuous
  2. within-subjects has at least 2 categorical variables
  3. between-group has at least 2 categorical variables
  4. no sig. outliers for between and within
  5. dep. normal distribution for each combo of groups
  6. needs to have homogeneity of variance for each combo of groups
  7. sphericity must be equal
75
Q

How do you posthoc a mixed ANOVA?

A

Use a Bonferonni correction

76
Q

Discuss why people are beginning to question p values and “classic” null hypothesis testing?

A

P-values no not indicate unreliability. They actually measure whether a result can be attributed to chance, not the actual odds of a hypothesis occurring. CIs are a better representation of uncertainty and are more likely to be interpreted correctly

77
Q

What is the “New Statistics”?

A

Promotes reporting of confidence intervals and effect sizes versus p-values.
Use estimate based research questions e.g. “to what extent does… or how large of an effect size does…” .
Full declaration of intended procedure and data analysis
Use meta-analytic thinking through out

78
Q

Explain the basic logic of Bayesian statistics and why some think it is more accurate than using null hypothesis testing.

A

Examines how much is a group different rather than is a group different. Bayesian stats provide complete information about parameters. B more intuitive than NHST. B can both accept and reject the null hypothesis and examines the quality of data through confidence intervals.

Stats 580 (1 decks)

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