Test Flashcards
What are the 4 principles of the BPS code of ethics?
Respect, competence, responsibility, integrity
Which ethical guidelines form part of the ‘Respect’ principle of the BPS code?
Informed consent, right to withdraw, confidentiality
Which ethical guidelines form part of the ‘Responsibility’ principle of the BPS code
Protection from Physical and Psychological harm, Debrief
Which ethical guideline forms part of the ‘Integrity’ principle of the BPS code
Deception
Name the 5 non-parametric tests on the specification, what are the conditions for the use of each one, and how should you compare the observed and critical value for each?
- Chi-Squared: Independent Measures Design, Nominal Data. Observed value must be greater than or equal to the critical value for findings to be significant.
- Mann-Whitney U Test: Independent Measures Design, Interval/Ordinal Data. Observed value must be less than or equal to the critical value for findings to be significant.
- Binominal Sign Test: Repeated Measures Design, Nominal Data Observed value must be less than or equal to the critical value for findings to be significant.
- Wilcoxon Test: Repeated Measures Design, Interval/Ordinal Data. Observed value must be less than or equal to the critical value for findings to be significant.
- Spearman’s Rho: Correlations. Observed value must be greater than or equal to the critical value for findings to be significant.
How do you carry out the Chi-Squared statistical test?
- Add one row and one column to the table and total all the rows and columns, also giving you an overall total, which should be equal to the number of individuals in the sample.
- Calculate expected value for each box in the table. Expected value = (Row Total x Column Total) / Overall Total
- Carry out the Chi Squared Formula: χ^2 = ∑[(observed value – expected value)
^2] / expected value - Calculate degrees of freedom: df= (number of rows - 1) x (number of columns - 1). Do not include the rows/columns you added for totals in this
- Find your calculated df in the Chi squared critical value table, and read to 0.05 unless specified otherwise. This will give you the critical value. If the observed Chi-Squared value is greater than or equal to the critical value, findings are significant.
How do you carry out the Mann-Whitney U test?
- Add a ranking column to the table for each condition, and add an extra row labelled “R1” in the first condition and “R2” in the second condition.
- Rank all of the data points, making sure to rank the data from both conditions together. Add the rankings to the table, and total them up for the R1 and R2 values.
- Whichever is the smallest value out of R1 and R2 should be used to carry out the formula: U1 = R1or2 - [n1(n1+1)/2]. This will give you your observed value.
- Use the Mann-Whitney critical value table to find the critical value. N1 is the number of participants in condition 1 and N2 is the number of participants in condition 2. If the observed value is equal to or less than the critical value, findings are significant.
How do you carry out the Binomial Sign test?
- Add a column to the table titled “Direction”. In this column, put a plus for yes-no, a minus for no-yes and “ignore” for yes-yes and no-no. The plus and minus can be switched around as long as it is consistent throughout the table.
- Total the amount of each sign. Your observed value is equal to the frequency of the least frequent sign.
- N is equal to the number of participants with a flow of direction, so add up how many participants there are who don’t have “Ignore”. Find this number of the critical value table, and read across to 0.05 unless specified otherwise. This will give you your critical value.
- If the observed value is less than the critical value, findings are significant.
How do you carry out the Wilcoxon Signed Ranks test?
- Add a column to the table labelled “Difference”. Find the difference between the 2 values.
- Add another column labelled “Ranked Difference”. Rank the differences, ignoring the positive and negative signs. Ignore any values that are 0.
- Total the amount of each sign, ignoring any 0s. Which sign is the least frequent, use the values in the table that have this sign.
- Add the ranks together of the values in the table with the least frequent sign. This is your observed value.
- Find critical value in Wilcoxon critical value table. N is number od differences, therefore amount of participants excluding any ignored 0 difference values.
- If the observed value is less than the critical value, it is significant.
How do you carry out Spearman’s Rho?
- Rank the values in the first co-variable. Also rank the values in the second co-variable, but ensure to rank the 2 co-variables separately.
- Create a table with value of co-variable 1, rank of co-variable 1, value of co-variable 2, rank of co-variable 2, difference between ranks and difference squared. Insert all data into the first 4 columns of this table, then calculate the final 2 columns.
- Add up all of the difference squared values. Carry out Spearman’s Rho formula: r = 1 - (6ΣD^2)/[n(n^2-1)]
- This will give your observed value, which is also the correlation coefficient
- Find critical value in Spearman’s Rho critical value table, read from number of pairs, across to 0.05 unless specified otherwise.
- If the observed value is more than or equal to the critical value, the findings are significant
Give a strength and weakness of using the mode
- Strength: Not skewed by extreme values
- Weakness: Doesn’t take all the scores in the data set into consideration
Give a strength and weakness of using the median
- Strength: Not skewed by extreme values
- Weakness: Doesn’t take all the scores in the data set into consideration
Give a strength and weakness of using the mean
- Strength: Takes all of the scores in the data set into consideration
- Weakness: Can be skewed by extreme values
How do you calculate the Range?
Calculate the largest value minus the smallest value
How do you calculate the Variance and standard deviation?
Subtract the mean from each number in your sample, and square the result of these calculations. Add these numbers together and divide this by how many numbers you have in the sample minus 1. This will give the variance. Square root the variance to find the standard deviation
