Quiz 1/Midterm 1 Flashcards
What are statistics?
Rules and mathematical procedures used to present results of investigations
What are the two broad classes of statistic?
Descriptive statistics and inferential statistics
What are descriptive statistics?
Describe observations/relationships (e.g. mean, range). Straight forward. Based on a sample.
What are inferential statistics?
They take observations and make conclusions beyond people you observe. They make generalizations about a population. Involves probability.
What is the scientific method?
A way of finding things out about the world. It is a combination of objective observation (descriptive stats) and reasoning/logic.
What is inductive logic?
Inductive logic goes from a specific premise, and makes a generalization to a general conclusion. Ex. My cat has fleas so all cats have fleas. Uses inferential stats.
What is deductive logic?
Goes from a general premise to specific conclusion, not using statistics. All cats have fleas, therefore my cat has fleas.
What is a variable?
An attribute/property/condition being observed.
What are the goals of the scientific method?
1 - Objectively observe and describe lawful relationships b/n variables (descriptive stats)
2 - Predict new observations/generalize using inductive logic.
3 - Explain why observed relationship occurs (Theory)
What is a constant?
An attribute that only takes on one value (ex. room full of people with same hair colour).
How many variables must be involved in a relationship?
2
What is sampling error?
The sample may not be representative of the population. A larger sample = less chance of error.
What is the difference b/n theory and an observation?
Theory gives a REASON for association
How do you test a theory?
Theory > Operational Definition > IV and DV > Experimental Hypothesis > Sample > Random Assignment > Make observations > Descriptive Stats > Inferential Stats > Leap back to theory
What is an operational definition?
Defines variables in terms of operation you use to measure it.
What is an experimental hypothesis?
Explaining what you expect to happen (NOT why)
What is a random sample?
Sample randomly chose from entire population (rare)
What is a convenience sample?
Uses sample that is available (ex. uni students)
What can bring about error in the scientific method?
Problems in operational definition (not measuring what is intended, confounds (extraneous/nuisance variables), and sampling error (atypical sample)
Why are the type of study/scale important?
Helps determine what type of statistical analysis, determines proper way to graph, and determines strength of conclusions.
What is a manipulated variable?
Experimentor controls, makes it one way, determines condition
What is a measured variable?
Observe and if possible, quantify/categorize
What is the nominal scale of measurement?
Simplest - categories. Qualitative.
What is the ordinal scale of measurement?
Ranking (e.g. 1st, 2nd, 3rd). Distances b/n intervals may not be equal. Can’t do averages.
What is the interval scale of measurement?
Rating scales, personality tests. Quantitative. Equal intervals. Can use averages.
What is the ratio scale of measurement?
Quantitative. Equal size intervals. True zero point. Can use ‘2x’.
What is a true experiment?
All IV are manipulated. Random assignment. DV measured. Testing a theory. Use a bar or line graph.
What is a correlational study?
Least control. All variables are measured - no manipulations, No IV or DV. Use correlation coefficient ‘r’. Use a scatter plot to show an association. One variable does not necessarily cause another.
What is a quasi-experiment?
One or more non-manipulated IV. Divide participants into groups based on previously existing conditions. Only measured variables.
What is the difference b/n real and apparent limits?
Real limits are in between intervals (eg 8.5-9.5), and apparent limits are values of scores that are highest and lowest in an interval (eg 9)
When are grouped frequency distributions used?
When you have a wide range in variable you are measuring.
What are grouped frequency distributions?
Frequency distributions made up of an ordered list of class intervals.
What is a class interval?
Lower limit - Upper limit
What are the rules for a professional looking frequency distribution?
1 - 5-15 Class intervals
2 - Convenient counting interval size
3 - All intervals equal size
4 - Lower limits have to be even multiples of chosen interval size
How do you make a frequency distribution?
1 - Calculate # of rows (Range+1)
2 - Divide by 10 to get interval size
3 - Choose convenient counting interval
4 - Make lowest apparent limit = multiple of interval size.
How many decimals of accuracy do we need?
2
What does a normal distribution look like?
Bell curve, symmetrical
What does a positive skew look like?
Tail to the right. Average is higher than most scores.
What does a negative skew look like?
Tail to the left. Average is lower than most scores.
What does central tendency tell you?
The typical score.
What is the mode?
Most common score. Can be no mode or more than one. Can also be used with nominal data.
What is variability?
Individual differences within the data. Variability determines whether measures of central tendency are accurate.
What are some measures of variability?
Range, standard deviation
What is the standard deviation?
Typical distance between each score and the mean. (Population /n, and sample /(n-1) )
What is the definitional formula for SS?
SS = Sum(X-M)^2
What is the variance?
SD^2. The squared typical distance between each point and the mean.
What are statistics?
Set of mathematical procedures for organizing, summarizing and interpreting information
What is a parameter?
Value of numbers that describes a population
What is a nonequivalent groups study?
Research in which collections of participants are formed without the researcher controlling assignment
What is a pre-post study?
Evaluation of a treatment’s effect by comparing observations before and after the treatment
What does biased mean?
Description of a statistic for which the average value is unequal to the parameter
What does unbiased mean?
Description of a statistic for which the average value is equal to the population parameter
What is a biased statistic?
Metric that consistently tends to over/underestimate the population parameter
What is an unbiased statistic?
Metric that provides an accurate estimate of the corresponding population parameter