6. Sign Test

• 1. Sign Test

  Used when looking for difference.

  • Used in repeated measures design.
  • Nominal data.
  • *Only statistical test that will be asked to calculate in exam!!!!!*
• 2. Sign Test

  STEP 1:

  • Record data.
• 3. Sign Test

  STEP 2:

  • Data needs to be turned into nominal data.
  • Done by saying whether participant score higher or lower in 2nd condition.
  • State the sign (+) or (-).
  • If participant has same score (as in no change), then data ignored (put zero).
• 4. Sign Test

  STEP 3:

  • Add up amount of + and -.
  • Take the lower/smallest number.
  • This smallest number = observed value, needs to be compared to statistical table.
  • The smallest number is quoted as S = ‘number’.
  • E.g. ‘S = 6’.
• 5. Sign Test

  STEP 4:

  • Using statistics table, find critical value.
  • N = total number of participants (ignoring anyone who scored same in both conditions).
  • N = 19.
• 6. Sign Test


  STEP 5:

  • Refer to hypothesis, is it directional (one-tailed) or non-directional (two-tailed).
  • Directional (one-tailed) = change predicted to occur in only one direction.
  • Non-directional (two-tailed) = doesn’t predict direction of difference/relationship.
• 7. Sign Test

  STEP 6:

  • Go down 0.05 column for either directional (one-tail) or non-directional (two-tail) test, depending on which one is correct for the question.
  • Go across at the correct N to find the critical value.
  • E.g. Critical value = 5.
  • Always use 0.05 column for correct tailed test… UNLESS it’s a drug trial, which uses 0.01.
• 8. Sign Test

  STEP 7:

  • Compare calculated/observed value to the critical value.
  • E.g. calculated/observed value, S = 6 and critical value, 5.
  • For the study to be significant, calculated/observed value must be less than/equal to the critical value.
  • If not, null has to be accepted.
  • In above example, hypothesis is significant.