Stats tests

Cards (24)

  • Why do we use Chi-Squared?
    • There is a difference in ….
    • It produces nominal data of ……
    • It has an experimental design of …….
  • How do you write a statement for Chi-Squared?
    x² = ______, df = ______, p > 0.05
  • How do you write a conclusion for Chi-Squared?
    • The results are/not significant
    • Because the observed value of _____ is greater/less than the critical value of _____ at a p value of 0.05 (95%). Although 5% could be due to extraneous variables
    • Therefore, you would accept the null/alternate hypothesis of ______ and reject the alternate/null
  • What is the criteria for parametric testing?
    • Populations drawn should be normally distributed
    • Variances populations should be approximately equal
    • Should have at least Interval or Ratio data
    • Should be no extreme scores
  • What is the criteria for non-parametric testing?
    • Looking for difference or relationship between variables
    • Ordinal, Interval or Nominal data
    • Independent or repeated measures
  • How do you calculate Chi Squared?
    1. Label columns A , B , C and D
    2. Total rows and total columns (corner is just total of the 2 rows)
    3. To work out e for each = total row x total column / total in corner
    4. Do O-e for each
    5. Square each o-e
    6. Divide each o-e squared by e
    7. Add all together
  • How is Spearman’s Rho calculated?
    1. Rank data and rank both sets separately (If 2 or more numbers are the same work out the mean of the rank)
    2. Find difference between ranks
    3. Square differences
    4. Use rest of formula
  • When is Spearman’s Rho used?
    • Relationship
    • When data is ordinal or interval
  • How do you write a statement for Spearman’s Rho?
    Rho =______ > n = _____ , p > 0.05
    To be significant the observed value must be equal or greater than the critical value
  • When is a Mann Whitney - U test used?
    • DV produces ordinal and interval data
    • Independent measures
    • Differences
  • How is a Mann Whitney - U test calculated?
    1. Rank data as one group (2 or more numbers in same rank find mean of rank)
    2. Total ranks of each group
    3. Substitute into formula
    4. Choose smallest answer
    5. Use no. of ppts from each group to find critical value
    6. Observed value must be less than or equal to critical value to be significant
  • How is a Mann Whitney - U statement written?
    U = (observed chosen) ; n1 = ; n2= , p < 0.05
  • How is a Wilcoxson Signed Rank calculated?
    1. Find difference between scores
    2. Rank the differences (ignore negatives)
    3. Sum of ranks of positive differences
    4. Sum of ranks of negative differences
    5. Choose smallest number
  • When is a Wilcoxon Signed Rank test used?
    • Difference
    • Repeated measures
    • Ordinal data
  • When is the research significant when using a Wilcoxon Signed Rank Test?
    • Observed value is less than or equal to the critical value
  • When is a binomial sign test used?
    • Difference
    • Repeated measures
    • Nominal data
  • When is research significant when using a binomial sign test?
    • Observed value is equal to or less than the critical value
  • What does P<0.05 mean?
    • The probability of the null hypothesis being true is less than 5%
    • Therefore there will be a difference in....
  • How is the Degree of Freedom calculated for Chi squared?
    (Row - 1) x (Column - 1)
  • How is a statement written for Wilcoxon Signed Rank?
    T = , n= , P < 0.05
  • How is a statement written for Binomial?
    S = , n= , p < 0.05
  • Label the distribution curves
    A) Negative
    B) Normal
    C) Positive
    D) Mean
    E) Median
    F) Mode
    G) Mean
    H) Median
    I) Mode
    J) Mode
    K) Median
    L) Mean
  • How is a variance conclusion written?
    • There is more / less variation between the (variables) because (data). Therefore this tells us (spread around mean) is more / less consistent. This suggests (go outside box)
  • Why is standard deviation used instead of the variance?
    • Variance is an untypical figure and standard deviation is more representative of the original data and a more realistic spread that is in line with the measure used