within-subjects design Flashcards
(40 cards)
Within-Subjects Design
When each participant in exposed to each and every experimental condition
deductive Model
1. Theory (previous literature or
Existing theory)
2. Hypothesis (are found in real
life)
3. Prediction (are found only in
my experiment)Extraneous Variable for within-subjects
Huge variability of individuals we perfectly match extraneous variables by using the same person in each condition
Within-Subjects biggest confound
The Order Effect: the order/timing of conditions cause systematic bias effects, testing effects and boredom effects
How do we fix The Order Effect?
By Counterbalancing, it equalises theses effects across conditions so they cancel each other out.
What is the factorial Effect?
( # of trials != all possible combinations)
i.e 4!=4x3x2x1
Method used when you need to counterbalance more than 2 conditions
Advantages and Disadvantages of Within-Subjects design
\+ Best control of potential confounds \+ Focus on differences between conditions \+ Need fewer participants - Order Effects are inevitable: - Contamination - Asymmetric Carry over - Time - Difficult to counter balance if their are more then 2 conditions of IV - Contrast effects.
What are the 4 Types of Validity?
1. Construct Validity:
How well do the IV and DV
measure our construct
2. Internal Validity:
How well does the experiment
design control for confounds
3. External Validity:
How well do our conclusions
generalise to real life
4. Statistical validity:
How well are the conclusions
supported by the experimental
dataType 1 error & Type 2 error
Type one error: Alpha
Is a false positive, when you reject the null hypothesis when the null hypothesis is true.
Type two error: Beta
False negative: fail to reject the null hypothesis when the null is wrong
Variance
Measures how spread out your data is
Standard Error of the Mean
Measures how close your sample mean is to the real mean
Independent t-test
t formula
t= Group dif. /standard error of sample
Independent t-test (between-subjects design)
t formula
t= Group dif. /standard error of sample
Independent t-test for
between-subjects design
t= group difference/ standard error of the mean
t is big if
- group difference is big i.e. homogeneous sample
- standard error of the mean is smaller i.e. bigger n
Paired t-test design for
(within-subject designs)
Types of hypothesis
non-direction or directional
p-value
the probability that your t result size is purely due to chance i.e. probability of type 1 error, a false positive (alpha) you reject the null hypothesis even though the null hypothesis is true.
p
Sample distribution
if t=0 there is no difference between groups
With random assignment we are not always gonna be equal variance thus, we use sample distribution. Because the larger your sample size the closer your sample mean will be to the real mean.
t-distribution
if my t is very big then it’s very unlikely that my two means are from the same population.
bigger the t-value the smaller the p-value
the more variance between groups means the smaller the probability that our results are due to chance
A t-value of 0 indicates that the sample results exactly equal the null hypothesis.
As the difference between the sample data and the null hypothesis increases, the absolute value of the t-value increases.
Sampling distribution
is where multiple t-values are calculated from the same sample and plotted as distribution graph.
if t- value is big we…
reject the null hypothesis. Because a large t-value suggests that there is large variance between both groups. But we know that on a t-distribution groups are form the sample sample split into two groups. Thus, we can assume that the variance between groups is caused by our IV.
Critical value of t
the critical value of t corresponds o a p-value where p
