Biological Psychology: Hypothesis Testing & Comparing Means

Flashcards covering hypothesis testing concepts from Psych 101, specifically focused on independent-measures and paired-samples t-tests, effect sizes, and statistical assumptions.

When is an independent-measures t test typically used?

It is used when comparing two groups exposed to different experimental conditions to see if they differ on a particular measure.

In the alcohol use disorder (AUD) reaction time study, what were the average reaction times for Group A (AUD) and Group B (no AUD)?

Group A (AUD) had an average reaction time of $600\,ms$ and Group B (no AUD) had an average reaction time of $500\,ms$.

What does the null hypothesis ($H_0$) state regarding the difference between two means in an independent-measures t test?

The $H_0$ states that the difference between the population means is zero.

Why is pooled variance calculated in an independent-measures t test?

It is used to combine the two sample variances into a single value when calculating the standard error of the mean difference.

In the provided reaction time example for independent measures, why was the result considered non-significant?

Because the empirical t-value ($tempirical = 1.873$) was less than the critical t-value ($tcritical = 2.447$ for two tails at $\alpha = .05$).

How does a paired-samples t-test differ from an independent-measures t-test in terms of participants?

A paired-samples t-test obtains two measures (such as treatment and baseline) from the same group of participants, whereas independent-measures uses two different groups.

What was the empirical t-value for Experiment 2 (the paired-samples reaction time study)?

The empirical t-value was $13.464$.

Why can a very small effect still be statistically significant in a t-test?

Because the t-test is highly dependent on sample size; a very large sample size ($n$) can result in statistical significance even for minor effects.

How does sample variance ($s^2$) affect the likelihood of obtaining a significant t-statistic?

A large sample variance produces a larger estimated standard error in the denominator, which results in a smaller t-value (closer to $0$), making a significant effect less likely.

What are the two effect size measures for t-tests discussed in the lecture?

The two measures are Cohen’s $d$ and the amount of variance explained, $r^2$.

What is the definition of Cohen's $d$?

Cohen's $d$ is an estimate of effect size independent of sample size, calculated as the mean difference divided by the standard deviation.

What are Cohen's (1988) conventional cut-off values for interpreting Cohen's $d$?

Small effect: $d = 0.2$; Medium effect: $d = 0.5$; Large effect: $d = 0.8$.

What are Cohen's (1988) guidelines for interpreting the effect size $r^2$?

Small effect: $r^2 \approx .01$; Medium effect: $r^2 \approx .09$; Large effect: $r^2 \approx .25$.

What is a Confidence Interval (CI) as defined in the notes?

An interval or range of values centered around a sample statistic (like a sample mean) that is likely to contain the corresponding population parameter.

What are the three mandatory assumptions for running t-tests?

1) Observations must be independent; 2) Populations from which samples are drawn must be normal (though t-tests are robust to violations with large samples); 3) For independent-measures tests, samples must have equal variances (homogeneity of variance).

According to APA7 rules, how should a p-value of .000 reported by JASP be written in a results section?

It should be reported as p < .001 because a p-value can never be exactly zero.

In statistical reporting, what is the rule regarding leading zeros for numbers that cannot be greater than 1 (like p-values)?

The leading zero should be omitted (e.g., write $p = .02$ instead of $p = 0.02$).