Week 1 Lecture, why do we need stupid stats? Flashcards
Why do we need statistics ?(3 )
to answer interesting questions, generate new knowledge and tells us odds of being wrong
What are the two types of data analysis and what do they test?(the two Q’s)
1) Quantitaive: testing theory’s using numbers
2) Qualitative: testing theory’s using language
What are the 5 steps to the research process? (IGGCA)
Initial observation
Generate a theory
Generate hypothesis
Collect data to test theory
Analyze data
What’s the difference between a theory and a hypothesis? geuss and ..
Hypothesis is a testable idea usually based on a theory (an assumption)
Theory is a principal or set of principals that explain known findings on a topic
What are the two ways you can conduct research and how? Finding something or someone finding..
Initial observation- find something needing explaining
Follow research stream: research developed over time by many researchers
What is a variable and it’s two types and what do they measure ?
anything that can be measured and can differ across entity’s in time
1) Independant Variable- what causes change
2) dependant Variable- what detects change (whats measured)
What’s a construct and two examples(something we can’t see) and what type of construct do we use instead?
Variable we are measuring indirectly. Variable cannot be measured or observed thus needs other measures to validify like a psychological construct that can be measured and defined.
Example: Anxiety,Depression
What is a research stream and how do you find it from an old study? literally like a stream
Stream of research is guided by a topic and like a stream bumps and ricks are in the way uncovering new questions. look backwards or forwards to see who identified it.
What does it mean by operational?
any type of variable that is able to be measured or manipulated
What’s falsification?
Disproving a theory/hypothesis
How do you write out The null hypothesis, what do the variables represent and what are the two names for research hypothesis?
NULL: H0: u1=u2
H-Null U1- Mean of population U2- effected
Non directional
Directional
What is a non directional research hypothesis and write it out
H1:x1 (bar)≠x2(bar)
Has no effect on variables
What is a directional hypothesis and write it out?
H1: x1(bar) > x2(bar)
Has an impact on one of the variables
Which of the two research hypothesis holds more power and why?
Directional hypothesis holds more power if you know the direction of change as the critical value shifts (value changes as one variable is being effected)
What are the 3 Names each for IV’s (PPM) and Dv’s (POM)
IV- Proposed cause, predictor variable, manipulated variable
DV- Proposed effect, outcome variable, measured not manipulated variable
Levels of measurement: NOIR + B What are they and one example for each
Nominal- no group is better than another: hair colour, gender
Ordinal- Set up by groups of hierarchy: Rate my professer (likerd scales), speed of cars
Interval-no set 0 point: Longitude and latitude, temperature
Ratio-has set 0 point: crime rate, age
Binary- In one group or another never both: morse code, dead or alive
What’s measurement error? and an example
Discrepancy between the actual value we are trying to measure and the number we receive as a result.
You weigh 100 lbs the scale says 103, 3 lbs is the measurement error
What are the 3 basic research designs? (CCE)
Correlational research
Cross sectional research
Experimental research
What is correlational and cross sectional research? (real world and different representerz)
Correlational- observing relationship between 2 variable without interfering
Cross sectional: data comes from different age groups and different people representing each
What is experimental research? (like partial correlation, what do we leave out 1,2…)
One or more variables systematically manipulated to see outcome on an outcome variable by itself or in combination
What’s the three criteria in Cause and effect (Hume,1748) Goes in circle almost
1) cause and effect must occur close in time together
2) cause must occur before the effect
3) no effect without presence of cause
What is another name for a confounding variable and what is it?
Tertium Quid
Variable we may not now or have measured, but still potentially has effect on outcome variable (DV)
( Methods of data collection) What are the 5 names for Independent variable (IBBUU) and Dependent variable (DRWPC)
Independent measure
between subjects design
between group
unpaired
uncorrelative
Dependent measure
repeated measures
within subject
paired
correlative
What are the three types of variation and what do they account for(1 example each)? (created by, created by, minimize)
- Systematic variation- differences in performance created by specific identifiable source (bias)
- Unsystematic variation- differences in performance created by unknown factors (age)
- Randomization- minimizes unsystematic variation
What is random selection? (defined) and random assignment? (groups) What is the difference between two?
Random selection: Choosing participant from defined population so you can state probability of individual being included
Random assignment : putting people in groups based off laws of random chance
-controls for variance we don’t know exists
What are Frequency distributions and what’s another name for them?
Histograms, Graph plotting values of observations on X axis with a bar showing how many times each value occurred in data set
What are the 2 property’s of Frequency distributions and explain what two kinds of things go into each property? (positive,negative tails and values) (who’s the short little blonde kid who worked at BB?) (platypus)
1) skew: symmetry of distribution
Positive skew: scores bunched at low values with tail pointing at high values
Negative skew: Scores bunched at high values with tails pointing to low values
2) Kurtosis -heaviness of tails
Platykurtic- Light tails (can be too flat)
Leptokurtic-heavy tails (can be too pointy)
Why is random sampling so hard? And why does the mean, mean nothing?
Hard to define a population for random sampling.
Mean has no measure of spread so the values can be all over the place (sprawled) or unseen and you wouldn’t even know
Why do we call stats rules assumptions and what happens if you break them?
Called assumptions as they are broken often and can cause a type 1 error (stating something untrue in experiment) When you do break an assumption do some research to see what to do next
What are the 3 measures of central tendency and how many of … does each contain? (MBM)
Mode-most frequent score 1 mode
Bi modal-2 modes
Multi Modal-7 modes
What is the normalicy assumption?
That most of our stats are based on a normal curve (bell curve), when we don’t use normal bell curve we violate normalicy.
How do you find the median mean and range?(middle, sum of scores, subtract)
Median- find middle number with data set lined up smallest to largest
Mean- find sim of scores and divide by numbers in data set
Range- subtract smallest score from the biggest
What’s the dispersion in Interquartile range and what of the median do they represent ? (3 quartiles)
1) second quartile- median
2) lower quartile -median of bottom half of data
3) upper quartile - median of upper half of data
What does variability measure?
Tells us how well the measure of central tendency portrays entire data set
What are two types of deviance and what does deviance do?
Deviance and total deviance
Calculates spread of scores by looking at how different each score is from the mean
What does IQR represent along with range?
IQR is range of values in middle of the score set and asses variability where most (middle) scores lie
Range: is dispersety of scores in data set, shows highest and lowest number scores go to
What does SEM and SD represent?
Sem- shows how well standard deviation could model real world
SD- show how dispersed data is in relation to the mean
What do COI represent?
Represent the range that will theoretically capture the true mean 95/100
REPRESENTED BY ERROR BARS