Karteikarten
Probability Distrubution
Karten in diesem Stapel (19)
- 1Probability DistrubutionCalculate probability of something occuring. Must have normal distrubution where mean =0 and DV=1. If it isnt we need to calculate z score. Z score calculated by taking data point minus mean and divide by s (SD). Then we compare z score with table of standard normal distrubution
- 2LimitsOnce we have z scores we can calculate limits of data set to see where middle 95% of scores fall
- 3What z scores is important for limits1.96=95%, 2.58=99%, 3.29=99.9% they determine top and bottom percent with rest falling in middel
- 4PopulationCollection of units to which we generalize our finding to. Can be large or narrow but very rare to access all members of populations so we use samples. Samples is a subset of the population.
- 5What is a sampleSample is subset of population to which we infer characteristics about the population. Bigger samples more likely its accurate. Results may vary slightly but will be similar
- 6What is sampling variations?When sample mean differs from Population mean and samples differ from each other
- 7Standard error of meanIf we take means of samples and calculate the overall average would give us the pop mean. So the SD will tell us how the sample mean represents the pop mean (Standard error) Small SE more likely to be accurate
- 8Confidence IntervalsBoundaries calculated in which true mean will fall. Large sample same as normal distribution limits (95%). SD=1 and mean =0. Upper boundary mean + (1.96xSE) and lower boundary mean - (1.96xSE)
- 9What does it mean when its significantly differentConfidence intervals does not overlap and we assume the means are from different population.
- 10Who discovered the null/alternative hypothesisNeyman and Pearson
- 11Explain null hypothesis and alternative hypothesisNull hypothesis is when there is no difference between population means (Bigger than pvalue of 0.05 null is retained). Alternative hypothesis is when there is difference between pop means (less than 0.05 its significant and reject 0)
- 12What is directional and non directional hypothesisDirectional hypothesis- predict the direction of the difference between population means. Non directional hypothesis does not predict direction.
- 13Hybrid & Fisher vs Neyman & Pearsonp value for null hypothesis. Determines whether alternative hypothesis is more likely to be true
- 14P valuereject null and significant
- 15One or two tailed testNon directional we use two tailed, directional we use 1 tailed.
- 16Type 1 errorWhen we conclude there is an effect present but in fact there isn't. Alpha level of 0.05 probability of occurring.
- 17Type 2 errorWe believe there is no effect present but there is. Probability of occuring is known as beta level. Max we use is .2. If we lower the Type 1 error the type 2 error likelihood will increase.
- 18Fit of modelDegree to which a statistical model represents data collected
- 19Bias in statisticsOutlier- Result that is different from rest of data, violations of assumptions- (parametrics) Additivity and linearty, homogeneity of variance, normality, independance
Informationen
- Beschreibung
- Bachelor of Psychology Evidence Based Approach Karteikarten am Probability Distrubution, erstellt von megandee am 02/05/2014.
- Erstellt von
- megandee
- Aufrufe
- 13
- Bewertung