RM statistics

17 flashcards · Other · by Daniel F.

1. What to do with missing values

   Listwise deletion → completely delete from the dataset, N changes Pairwise deletion → participants still included, missing answers removed Replace missing values by the mean value → manipulating results but might be necessary when little amount of participants

2. Reliability analysis

   cronbach's alpha, only when all questions interval/ratio data type When value higher than 0.6 reliability sufficient

3. Can you always sum variables

   no, only interval/ratio data types

4. Choosing statistical techniques to test hypothesis

   How many variables are involved What is the data type of the involved variable Asymmetric vs symmetric

5. How many variables are involved

   One: Univariate analysis (descriptive statistics) → mean number, how widely ppl differ Two: Bivariate analysis (inferential statistics

6. What is the data type

   Nominal Ordinal interval/ratio

7. Asymmetric vs symmetric

   Asymmetric when: Variables have different data type OR predicting DV from IV (causal relationship) Else: symmetric

8. Univariate statistical techniques

9. Bivariate statistical techniques - symmetric

10. Bivariate statistical techniques - asymmetric

    k = 2 two groups (male or female) k > 2 more than two groups (e.g. 5 religions) X: IV Y: DV

11. Related vs. unrelated sample

    Related sample (within-subjects design) has each participant in each condition, small sample needed and control for intra-individual effects Unrelated sample (between subjects design) has each participant in only one condition and no practice effects or hypothesis guessing always assume

12. Aim of statistical tests

    provide information about the generalizability of the results found in a study (from sample to population)

13. Statistical test 3 parts

    Test statistic: number which gives information about the test outcome P-value: probability that the outcome is based on chance Degrees of freedom (df): number of independent observations based on which the test statistic is calculated (not applicable to all tests) If degrees of freedom increases, likelihood of p-value will decrease

14. How to write a hypothesis

    Motivation and satisfaction are positively related The more motivation, the more satisfaction, vice versa H0: r = 0 → H0 is always the hypothesis of no-effect H1: r > 0

15. interpret p-value

    suppose the null-hypothesis is true, how big is the probability that you would find this particular result? A really small p-value indicates that the null-hypothesis not is likely to be true Typical cut-off point = alpha = 0.05 P <= 0.05: we reject the null hypothesis (so we believe the result is real and meaningful) P > 0.05: we do not reject the null hypothesis (so we do not believe the result to be real and meaningful - it is due to chance)

16. which statistical test does what

17. How to report results

    A. What is the goal? B. How did you do that? C. What was the result? D. Which conclusion?