1/81
Looks like no tags are added yet.
Quantitative data
Data expressed as numerical values, such as age in years.
Qualitative data
Data expressed as categories or verbal descriptions, such as adult versus juvenile.
Surrounding environment
How we perceive an element depends on the context or objects around it.
Visual crowding
Nearby visual elements can interfere with recognizing an individual element.
Law of similarity
The mind groups elements similar in form, color, size, or brightness.
Law of proximity
Elements close together in space or time tend to be perceived as a group.
Subjective color perception
Color experiences differ across people, so color choices affect interpretation and accessibility.
Aesthetic preferences
People have visual preferences; attractive graphs may be trusted more.
Encoding object
Visual mark used in a graph, such as a bar, line, dot, or wedge
Value-encoding attribute
Feature that carries a numerical value, such as bar height, dot position, or color
Pie chart
Shows proportions of mutually exclusive categories that together total 100%
Bar plot
Compares values across separate, usually categorical, groups using bar height
Line graph
Connects values across a continuous quantitative x-axis
Scatter plot
Shows relationship between two quantitative values. May require a legend. May have a best fit line
Histogram
bar plots specifically that plot frequency (the nuumber of times a certain value occurs in the data)
Bin width
Size of each histogram interval; changing it can change the apparent distribution shape
Heatmap
Uses color to encode values in a matrix of row-and-column combinations
Error bar
Visual indicator of variability or uncertainty
Standard deviation
Measure of how spread out individual values are around their mean
What are ways to ensure chart clarity?
Use color wisely and avoid having too much or too little information
Inconsistent axis spacing
Using unequal numerical intervals as if they were equally spaced can distort trends
Truncated axis
Axis that starts above zero or omits part of its range; can exaggerate visual differences
Inverted axis
Axis with values displayed in reverse order; can change how viewers perceive a trend
System 1 thinking
Fast, automatic, largely unconscious thinking influenced by heuristics and experience
System 2 thinking
Slow, controlled, conscious, logical, and systematic thinking
Heuristic-based bias
Bias caused when an otherwise useful shortcut is misapplied, especially under time pressure
Error-management bias
Bias shaped by unequal costs of mistakes; one type of error may be safer than another.
Artifact in bias research
Apparent bias arising from artificial testing or how researchers frame a problem
Gambler's fallacy
Mistaken belief that previous independent outcomes change the probability of the next one
Anchoring
Relying too strongly on an initial reference point when estimating or deciding
Decoy effect
Adding a less attractive third option changes preferences between two original options
Framing effect
judgment changes depending on how equivalent information is presented
Absolute versus relative risk framing
A change may sound larger when reported as a relative percentage than as an absolute difference
Sunk-cost fallacy
Continuing an activity because of past time, effort, or money despite poor future benefits
Availability heuristic
Judging frequency or likelihood by how easily examples come to mind
Confirmation bias
Seeking or favoring information that supports existing beliefs while discounting challenges
Examples of how to prevent cognitive biases:
develop awareness of biases, go through training, minimize time pressure
Randomness intuition
People often expect random sequences to alternate more and contain fewer streaks than they do
Alternation rate
Probability that the next item in a two-outcome sequence differs from the previous one
Hot-hand phenomenon
A streak of success suggesting success may continue; such claims require comparison with chance
Pareidolia
Tendency to interpret an unclear (or random) visual stimulus as
something familiar
Coincidence
a remarkable concurrence of events or circumstances without apparent causal connection
Curated coincidence
Coincidental event has been selected for some salient property from a large number of events
Staged coincidence
An apparently surprising concurrence deliberately arranged to occur
Correlation
Association between measurements, which does not by itself establish causation
Correlation coefficient
captures the extent to which two values are linearly related (-1 to +1)
False-cause fallacy
Assuming that an association proves one event caused another
Post hoc ergo propter hoc fallacy
Assuming an earlier event caused a later event simply because it came first
what is the Data-literate approach to coincidence
Assess chance, look for curation or staging, gather data, evaluate causation, then update beliefs
Probability
Number from 0 to 1 representing how likely an event is; 0=impossible, 1=certain
Addition Rule in Probability
the probability that one or both events occur
mutually exclusive: P(A or B) = P(A) + P(B)
not mutally exclusive: P(A or B) = P(A) + P(B) - P(A and B)
Mutually exclusive events
Events that cannot occur together (independent)
Multiplication rule in probability
the probability that both events occur together
P(A and B) = P(A) * P(B|A)
independent events: P(A and B) = P(A) * P(B)
Conditional probability
Probability of A given B, written P(A|B)
Prior probability
How plausible a hypothesis is before considering new evidence
Likelihood
Probability of observing particular evidence if a hypothesis is true
Posterior probability
Updated probability of a hypothesis after considering evidence
Bayes' rule
P(A|B) = [P(B|A) × P(A)] / P(B)
Base rate
Overall prevalence or probability of an event in the relevant population
Base-rate neglect
Ignoring overall prevalence when interpreting evidence about an individual case
Survey
Questions asked of a sample to make inferences about a larger population
Opinion poll
Survey designed to measure views about an issue, policy, or candidate
Sampling variability
Differences between estimates that occur when different samples are selected
Statistical (sampling) error
Random error caused by using a limited sample; generally decreases with larger random samples
Systematic error
Bias caused by flawed sampling or design; making a sample larger does not remove it
Law of large numbers
As a random sample grows, its average tends to approach the true expected value
Representative sample
sample that adequately reflects the population to which findings will be generalized
Probability sampling
Sampling approach in which selection is random
Random selection
Choosing participants randomly to help avoid selection bias
Selection bias
Systematic distortion when some groups are more likely than others to enter the sample
Nonresponse bias
Distortion when people who do not respond differ systematically from respondents
Convenience sample
Sample chosen because people are easy to reach
Weighting
adjusting how much different respondents count to better reflect population groups
Oversampling
Intentionally recruiting extra participants from a group, then adjusting its analytic weight
Margin of error
Half the width of the lecture’s 95% interval; smaller with larger samples
95% confidence interval
a range of values that you can be 95% sure that it has the true mean
Leading question
Question phrased to make one response more likely
Double-barreled question
Question asking about two issues at once, making one answer ambiguous
Acquiescence bias
Tendency to agree with survey statements
Question-order effect
Earlier questions influence responses to questions that follow
WEIRD sample
Sample dominated by Western, Educated, Industrialized, Rich, Democratic populations
External validity
Extent to which study findings can be generalized to people beyond those sampled