Single-Variable, Independent-Groups Designs + The normal distribution + t-test for unpaired observations + Flashcards

1
Q

What is the ex post facto design?

A

the researcher observes currently present behavior to then relate it to past experiences. There is no control for confounders and no causality can be determined , meaning rival hypotheses can not be rejected.

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

one group, single post-test design

A

there is a manipulation, but since there is no pre-measurement, there are still confounding factors and no causality can be established.
There is no control over possible placebo-effects, maturation, history, or regression to the mean.

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

One group, pre- and post-test design:

A

Nonexperimental
design in which participants are measured on a dependent
variable both before and after the manipulation., it is a time-series design.

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

Pre- and post-test, natural control group design

A

Design
in which preexisting groups are measured before and
after the manipulation of an independent variable.

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

How are experimental designs distinguished from non-experimental design?

A

Two elements: Control Groups or control conditions and random assignment.

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

Randomized, single post-test, control group design:

A

Experimental
design in which participants are randomly
assigned to two groups, and each group is tested on
the dependent variable after the independent variable
manipulation

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

Randomized, pre- and post-test, control group design:

A

Experimental design in which participants are randomly
assigned to two groups, and each participant is
tested on the dependent variable both before and after
the manipulation.

Recall that the pretest-posttest, natural control-group design discussed earlier is not
an experiment because it does not include random assignment to groups.

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

What can be some disadvantage of the pre-measurement

How can we try to control it?

A

pretest-manipulation interaction effect

Solomon’s four-group design
(Solomon, 1949). It is a powerful design that combines the randomized, pretest-posttest,
control-group design and the posttest-only, control-group design. However, because it requires the
resources of two experiments, it is not recommended for routine use.

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

What is unbiased assignment and how do we achieve it?

A

Assigning participants to groups
or conditions in such a way that the groups are statistically
equivalent at the start of the study. The most
common unbiased assignment method is random
assignment.

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

Multilevel, completely randomized between subject design:

A

A design with more than two groups, in
which each participant is randomly assigned to one
group. If there is a pre-test depends on the nature of the study.

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

What is a between-subjects design?

A

Different parrticipants in each condition

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

What are within-sunject designs or matched-subject designs?

A

the same participants or matched participants in each condition

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

What is ment by single-variable designs:

A

Designs that include just one independent

variable.

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

What would be the counterpart to single-variable designs?

A

factorial design

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

What characteristics has to have a statistics?

A
  1. The value can be calculated from the data

2. The probability distribution under H0is known

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

Name characteristics for unpaired t-test!

A

assume that analysed data is from a normal distribution.
there is also the assumption that the standard deviation (SD)s is approximately the same in both groups.
data of different participants in different conditions
(groups)
e.g. in an between subject design

17
Q

Characteristics of a paired t-test!

A

same as unpaired t-test but (!) : data of the same participants in different conditions

18
Q

How can we measure an unpaired t-test, if we have the data (means andstandard deviation) of two groups?

A

t=[(m1-m2)/(s1+s2)]x(-^n1+n29)

m1 & m2 are group means
s1 and s2 are standard deviations

19
Q

When is t more extreme and what does it mean?

A

When it deviates a lot from 0.

  1. if the effect (m1-m2) is more extreme (“signal”)
  2. the spread between groups (s1+s2) is smaller (“noise”)
  3. the number of subjects per group is larger
20
Q

What determines t?

A

the trinity of lecture 5 and 6… :D

21
Q

What does t has to be to have more evidence against the null hypothesis?

A

more extreme (high or low)

22
Q

Are the probability distributions of the t-test und H0 and that of the standard normal distribution (z) the same)

A

Only if n1 and n2 are extremely large

asthey are smaller the difference between t and z is bigger

23
Q

two more things that one should know about t distributions!

A
  1. they are a family of distributions
  2. the exact shape of t-distribution is determined by the “degree of freedom” The degrees of freedom refers to the number of independent observations in a set of data.
24
Q

At which t-value is the two tailed 5%p-value reached in the (1) t-distribution and in the (2) (z)distribution

A

(z) +- 1,96
(t) +- 2,57

now that matters!

25
Q

Can you please explain the difference between descriptive and inferential statistics?

A

Descriptive statistics: what are the main characteristics of the data? m1, m2, s1, s2.

Inferential statistics: what conclusions do the data allow about the population? t(18) = -3.24, p < 0.05.

26
Q

What is the advantage of random assignment?

A

Effective control for: history, maturation, regression to the mean, testing, and instrumentation.