Flashcards
Descriptive Statistics
Flashcards do conjunto (24)
- 1Types of reliabilty1. Test Retest- Measure same indvdual at two dfferent times. 2. Internal consistency reliability- Uses response at only one time and focus on consistency of items (Split half consitency and cronbach alpha) 3. Interrelated consitency- Evidence of reliability when multiple raters agree in their observation of someting
- 2Name 4 different types of validityFace, Content, Criterion related and construct
- 3Name two different construct validities and what they containConvergent validity- Same construct, related constructs, Discriminant validity- different construct, unrelated construct
- 4What is reactivity of measures?A measure is reactive when individuals changes their behaviour because they are aware of being measured.
- 5What must happen to a variable?Must be operationally defined (Method used to manipulate/measure the variable)
- 6What is level of measurement?Relationship between what is being measured and the numbers that represents what is being measured
- 7Name and explain the Level of measurement categoriesLevel of measurement is divided into Categorical variables (Name distinct entities) and continuous variables. The categorical vraiables are either binary (Variable divided into 1 0f 2 cat), nominal( Variables with equavilance are given specefic name or number) or ordinal ( Specifies order but does not specify what the difference in value is no absolute 0 point. Continuous variables are either discrete (count) or continuous (measure). This can be divided into interval (Equal intervals on scale represents equal difference) and ratio (O point , value)
- 8What is descriptive statistics?Describes the characteristics of data. Distribution, central tendency and dispersion are the main things we want to describe.
- 9Explain Dispersion?Counts how many times a score occurs. (Histogram) Value on x number on y. All shapes and sizes but ideally would by symmetrically distributed around centre
- 10Name and explain 4 distribution types.Raw scores (unmodified account of test performance), Distrubution (Set of test scores), Frequency distribution (Tally number of occurrences of test scores) and Grouped frequency distribution (Introduces intervals)
- 11Name three characteristics of a normally distrubutionUnimodal, Bell shaped, Symmetrical
- 12Deviation from normalityWhen its not symmetrical it is skewed. When more frequent at one end of scale at tails off toward the other. Positive and negative
- 13KurtosisDegree at which tails cluster at end of distribution (Positive-Leptokurtic and negative kurtosis- Platykurtic)
- 14Central tendencyLooks at the middle of data- Three tools mode, mean and median.
- 15ModeScore that occurs the most and thus its the tallest bar.
- 16MedianMiddel score when scores are ranked (n+1/2) If this falls between two number you sum up those two numbers and divide by two
- 17MeanMean is average- sum up all scored divide by how many scored there is . Create a model to compare the rest of our data to
- 18DispersionVariability or spread of scores in data. 5 dispersion types: Range, Interquartile range, Sum of squared errors (SS), Variance (S2), and Standrad Deviation (s)
- 19RangeSubtract Largest score with smallest score (can be effected by extreme scores)
- 20Interquartile rangeDivide set equally into 4 using 3 values. Subtract highest value with lowest value.
- 21DevianceDetermine the difference between our data point and the model. Data point minus mean. If - result it overestimates. The see the accuracy of the overall data set we can sum up the total error by summing up all the deviances.
- 22Sum of squared errorsIf we calculate the total errors the plus and minus will cancel each other out and result in 0, therefore we square each sum of error (SS)
- 23VarianceBecause so much sums and big sums in sum of squared errors we calculate the variance. The SS divided by n-1
- 24Standard deviationGet back to original units and not squared we square root the variance. SD tells us how our mean represents data. Small SD close to mean, Big is distant
Informações
- Descrição
- Bachelor of Psychology Evidence Based Approach FlashCards sobre Descriptive Statistics, criado por megandee em 02-05-2014.
- Criado por
- megandee
- Visualizações
- 19
- Avaliação