Labs n stuff
No description yet. This curated deck is ready for daily study.
Card preview (14)
Categorical vs continuous variables
Categorical: 2+ categories - Gender, ethnicity, multilingual vs monolingual, etc - Can also include numerical RANGES (e.g. 1-3 hrs, 4-8 hrs, etc – but not indivi numbers) Continuous: “infinite” # of values - Must have 7+ numerical values to be considered continuous enough - Age, height, GPA, test scores - # of hrs you spend doing “X” per wk
What’s wrong with this survey question: “On a scale from 1 to 7, how well do you sleep?”
Need to define the endpoints of the scale (e.g. 1 = not well, 7 = well)
What’s wrong with this survey question: “How often do you party?”
Vague – must define what constitutes “partying” + time interval (e.g. how many days a week do you drink more than 1-2 alcoholic beverages?) - Also need to clarify to round to nearest whole #
What’s wrong with this survey question: “How much do you disagree with the statement: I hate bananas”?
Has a double negative
What’s wrong with this survey question: “Rate the quality of your sleep and study schedules”?
Double-barreled question – should break into 2 questions, one for each var
In the Muller-Lyer illusion – Point of Subjective Equality vs Point of Objective Equality
Point of Subjective Equality (PSE): where the lines APPEAR equal (50% of the time I see it as longer, 50% of the time I see it as shorter) Point of Objective Equality (POE): the lines are ACTUALLY equal
Muller-Lyer illusion – how is the size of your illusion calculated?
PSE-POE = size of illusion Ex: if a person’s PSE is 182 pixels and their POE is 160 pixels, 182-160 = 22 pixels = size of their illusion
How does size of illusion relate to the Carpentered World Hypothesis
Carpentered World Hypothesis: your experience/exposure to corners plays a role in how strongly you see the illusion –> if you see a lot of corners, you’ll have a BIGGER illusion size –> if you’ve never seen corners, you’ll have a SMALLER illusion size
Limitations
• Characteristics of the study that may limit ability to draw conclusions • Specific for WHY it’s a limitation and HOW it might impact results – not generic (see below) • There will always be limitations – not necessarily an indicator of inaccurate results IMPORTANT: Good limitations are not generic - Correlation can’t infer causation, lab tasks may not reflect real world - Small sample size, convenience sampling, demand characteristics or experimenter bias
What are the 3 parts of a good limitation
1) Parameters: WHAT the researchers did that might be limiting 2) Justification: WHY it’s an issue 3) Prediction: HOW it might impact the study’s results
Ramirez & Beilock 2011 findings
Found that writing about testing worries right before a test boosts exam performance in classroom
Richland et al. (2009) findings
Found that pretesting (attempting to recall info before actually learning it) improved learning (as indicated by better test score)
Counterbalancing – progressive error/order effects
Practice effects (improved performance after repeated exposure), fatigue Ex: if I’m asked to eat a shitload of cupcakes and rate them, the last ones prob won’t taste as good as the first ones bc I’ll just be tired of eating cupcakes at the end
Counterbalancing – item effects
Certain items may be more memorable than others Ex: I’m asked to memorize words from 2 lists, but one list has words abt baseball while the other is just random words – obv I’d remember more words from the list w baseball words in it