Collecting Data

It limits inferential capability to confirm causality despite establishing temporal precedence and controlling for confounding variables

It enhances causative determinations by exclusively attributing variations to the treatment effect and inherent randomness in the process

It precludes the necessity of considering external causal factors given the sufficiency of internal validity provided by random assignment alone

It ensures the complete removal of subjectivity in decision-making, leading to pristine, objectified conclusions regarding cause-effect relationships

Larger p-value inaccuracies are detected as larger deviations from the normal distribution occur in insignificant ways.

P-value discrepancies are irrelevant as long as non-normal distributions sampled resemble a bell-shaped curve closely enough to pass muster.

Smaller p-value inaccuracies are detected as smaller deviations from the normal distribution occur in significant ways.

No discernible pattern emerges in terms of accuracy, as measured through the prism of deviation from the norms set forth at the beginning phase of experimental manipulation.

To estimate the range within which a population parameter lies based on sample data.

To calculate the exact value of a population parameter from sample data.

To prove that one categorical variable has an effect on another.

To determine if there is a statistically significant difference between two sample means.

The total variance for all samples combined must equal twice that of each separate sample's variance.

Heterogeneity of variances, meaning both populations have different but known variances.

Variance equality within samples rather than across populations being compared.

Homogeneity of variances, meaning both populations have equal variances.

Random assignment to treatment groups

Matching participants on specific characteristics

Conducting the experiment in a controlled environment

Using a large sample size for the experiment

The margin of error decreases and the statistical power increases.

The margin of error increases and the statistical power decreases.

Both margin of error and statistical power decrease.

Both margin of error and statistical power increase.

Convenience sampling

Stratified random sampling

Simple random sampling

Cluster sampling

Type I error rate increases while widths of corresponding intervals decrease.

Both Type I error rate and widths remain unchanged since α only affects p-values calculation.

Type I error rate decreases while widths of corresponding intervals increase.

Type I Error rate increases but has no effect on widths which are affected by other factors.

P-value from chi-square test for independence between predictor and response variables

Sensitivity or true positive rate

Positive predictive value

Specificity or true negative rate

The differences should be approximately normally distributed.

The variances within each group should be similar.

There should be no outliers in either group.

The sample sizes for both groups must be equal.