6-Significance Testing,Sampling Distributions & Z- Scores

0.0(0)
Studied by 0 people
call kaiCall Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/14

encourage image

There's no tags or description

Looks like no tags are added yet.

15 Terms

1
New cards
Descriptive vs. Inferential statistics
Descriptive statistics only summarise the data from your study sample (using central tendency and spread). Inferential statistics go beyond the sample to tell us how confident we are that an effect or difference actually exists in the wider population.
2
New cards
Three things that make us more confident a sample difference exists in the population
1) A larger difference between groups in the sample, 2) Smaller variation (spread/dispersion) within the sample, and 3) A larger sample size.
3
New cards
Null Hypothesis (H0) vs. Alternative Hypothesis (H1)
The Null Hypothesis (H0) states there is NO difference, association, or effect in the population (e.g., Power pose = Neutral pose). The Alternative Hypothesis (H1) predicts there IS a difference, association, or effect (e.g., Power pose ≠ Neutral pose).
4
New cards
Null Hypothesis Significance Testing (NHST)
A statistical method that assumes the Null Hypothesis (H0: no effect) is true and calculates how likely or surprising it would be to get our sample result purely by chance. If the result is very unlikely (p < .05), we reject H0.
5
New cards
What is a p-value?
The probability of obtaining your observed result (or one more extreme) purely by chance, ASSUMING the null hypothesis (H0) is true. For example, p = .0254 means a 2.54% chance of getting those results if H0 is true.
6
New cards
Common misinterpretation of p-values (ASA Statement)
A p-value does NOT tell you the probability that your alternative hypothesis (H1) is true, nor the probability that the null hypothesis is true—it only tells you how likely the data are assuming H0 is true.
7
New cards
Alpha level (α) & Statistical Significance (p < .05 vs. p > .05)
The alpha level (α) is the threshold for statistical significance, usually set at 5% (.05) in psychology. If p < .05, the result is statistically significant and we reject H0. If p > .05, the result is non-significant and we fail to reject H0 (accepting a 1 in 20 false-positive risk).
8
New cards
One-tailed (directional) vs. Two-tailed (non-directional) hypotheses
A one-tailed hypothesis predicts a specific direction (e.g., dogs are larger than cats; 5% rejection region in one tail). A two-tailed hypothesis predicts a difference without specifying direction (e.g., dogs and cats differ in size; 5% split into 2.5% in each tail).
9
New cards
P-value vs. Effect Size (Cohen's d)
A p-value only tells you how surprising a result is (statistical significance) and depends heavily on sample size. Effect size (e.g., Cohen's d: small = 0.2, medium = 0.5, large = 0.8) tells you the actual size/magnitude of the difference.
10
New cards
Sampling variability
The natural fact that every time you draw a different random sample from the same population, the observations and sample statistics (such as the sample mean) will be slightly different.
11
New cards
Sampling distribution of the mean & Expected value
A theoretical normal distribution showing the sample means you would get if you drew an infinite number of samples of the same size from a population. Its mean is called the "expected value" and equals the true population mean.
12
New cards
Central Limit Theorem (CLT)
States that the sampling distribution of the mean will ALWAYS follow a normal (bell-shaped) distribution if the sample size is large enough—even if the raw scores in the population are not normally distributed.
13
New cards
Standard Error (SE) & how sample size affects it
The standard deviation of the sampling distribution of the mean, estimated from a sample as SE = SD / √n. Larger sample sizes produce a narrower sampling distribution and a smaller SE (more accurate estimate of the population mean).
14
New cards
95% Confidence Interval (CI) & where 1.96 comes from
Calculated as Sample Mean ± (1.96 × SE), giving an interval with a 95% probability of containing the true population mean. The number 1.96 comes from the normal distribution, where 95% of all values fall within ± 1.96 SD of the mean (leaving 2.5% in each tail).
15
New cards
Standard Normal Distribution (Z-distribution) & Z-scores
A normal distribution standardised to have a mean of 0 and a standard deviation of 1. A Z-score, calculated as z = (score − mean) / SD, tells you how many SDs a score is from the mean and allows you to compare scores across different scales (e.g., IQ vs. exam %).