STATISTIC

• Independent variable


  The variable that is manipulated or changed to see its effect on the dependent variable
• Treatment conditions


  The different conditions or groups that the independent variable is tested on
• Between-subjects design


  An experimental design where each participant is only exposed to one level of the independent variable
• Within-subjects design


  An experimental design where each participant is exposed to all levels of the independent variable
• Matched


  Participants in different groups are paired up based on certain characteristics
• Dependent variable


  The variable that is measured or observed to see the effect of the independent variable
• 1. test for independent groups

  • Level of measurement is interval or ratio
  • There is one independent variable with two treatment conditions
  • The research design is between-subjects and participants are not matched
  • Used to determine if 2 groups are significantly different from each other on the variable of interest
  • Collects data from two separate samples and are normally distributed
  • Must have more than 5 values (data) from each group
  • The participants must be randomly sampled
• 1. test for matched groups

  • Level of measurement is interval or ratio
  • There are two independent variables
  • The research design is within-subjects and participants are matched
  • Used to test whether the mean difference between pairs of measurements is zero or not
  • Data from two groups are presented in pairs and are normally distributed
  • Sample size is minimum of two pairs
  • The participants must be randomly sampled
• One-way ANOVA


  • Level of measurement is interval or ratio
  • There is one independent variable
  • There are more than two treatment conditions
  • The research design is between-subjects
  • Compares variation between and within groups
  • Used to determine whether three or more group means are different when the participants are in the same group
  • When there is one categorical and one quantitative variable
  • Each subject only appears in one group
  • The variations within the groups must be similar for every group
  • The data must be normally distributed
• One-way ANOVA (repeated measures)

  • Level of measurement is interval or ratio
  • There are two independent variables
  • There are more than two treatment conditions
  • The research design is within-subjects
  • Used to determine whether three or more group means are different when the participants are in the same group
  • Each subject appears in each group
  • The data must be normally distributed
• Two-way ANOVA


  • Level of measurement is interval or ratio
  • There are more than two independent variables
  • There are more than two treatment conditions
  • The research design is a factorial design with multiple independent groups
  • Used to know how two combined independent variables affect a dependent variable
  • Independent variables should not be dependent on one another
  • Dependent variable should be normally distributed
• Two-way ANOVA (repeated measures)


  • Level of measurement is interval or ratio
  • There are more than two independent variables
  • There are more than two treatment conditions
  • The research design is a factorial within-subjects design
  • Used to understand if there is an interaction between the two factors in the dependent variable
  • Same subjects undergo both conditions
• Two-way ANOVA (mixed)


  • Level of measurement is interval or ratio
  • There are more than two independent variables
  • There are more than two treatment conditions
  • The research design is a factorial within and between-subjects design
  • Can be used when there are data of multiple levels of independent variable
  • Different subjects undergo each condition
• Mann-Whitney U-test


  • Level of measurement is ordinal
  • There is one independent variable with two treatment conditions
  • The research design is between-subjects
  • Used to compare outcomes between two independent groups
  • If the outcome is not normally distributed (skewed)
• Wilcoxon test


  • Level of measurement is ordinal
  • There is one independent variable with two treatment conditions
  • The research design is within-subject
  • Used to determine if 2 measurements from a single group are significantly different from each other on the variable of interest
  • If the data is not normally distributed (skewed)
• Kruskal Wallis test


  • Level of measurement is ordinal
  • There is one independent variable
  • There are two or more independent groups
  • Used if there are three or more categorical independent groups
  • If there are many different groups for every variable
  • Should be used if the variable is continuous or discrete
  • If the data is skewed
• Friedman test


  • Level of measurement is ordinal
  • There is one independent variable
  • There are three or more dependent groups
  • The research design is within-subjects matched groups
  • If the data is skewed
• Levels of measurement


  • Nominal
  • Ordinal
  • Interval
  • Ratio
• Nominal


  Category by labelling them, qualitative, discrete data and has no quantitative value, data only falls in one category only
• Ordinal


  Can be ranked in order but can't determine the interval between them, not a standardized scale, discrete data only, has no true zero
• Interval


  Determines the equal distance from one data to another, the data is quantitative and continuous, has a true zero
• Ratio


  Highest level of measurement, has quantitative and continuous data, has an absolute zero, all points have an equal amount of difference