08.+Working+Memory+and+Encoding

March 6, 2015
 * 08. Working memory and encoding**

Working memory Encoding long term memories (LTMs)
 * Outline**
 * Baddeley's model
 * Capacity of working memory
 * Intelligence
 * Depth of processing
 * Transcription versus understanding

Modal Model Summary: Working Memory
 * Working Memory**
 * Baddeley's Working Memory Model**
 * Short term memory - emphasizes memory storage
 * Working memory - emphasizes the //function// of working memory
 * Baddeley's Model consists of:
 * Central executive
 * "Slave systems"
 * Visuospatial buffer
 * Articulatory rehearsal loop
 * Central executive: the "boss" of working memory
 * Manages the slave systems
 * Initiates retrieval from long-term memory
 * Plans cognitive tasks
 * Initiates decision processes
 * Allocates attention
 * Maintains attention
 * Failures of central executive?
 * Mindwandering
 * Action slips: producing a habitual response that is inappropriate.
 * Articulatory rehearsal loop: temporarily maintains acoustic information
 * Evidence for articulatory rehearsal loop:
 * Articulatory suppression: repeat a very simple word ("la") over and over while learning word list. Interferes with short-term memory of the list
 * Word length effect
 * Baddeley et al. (1975): one vs. many syllables
 * Short: stoat, mumps, Greece, zinc
 * Long: rhinoceros, diphtheria, Australia, uranium
 * Results: same number of words, but they recall fewer when there are more syllables
 * Many IQ tests use digit span tasks (repeat back a list of digits). So are IQ scores higher in languages with shorter number words?
 * Naveh-Benjamin & Ayres (1986): Tested native speakers of English, Spanish, Hebrew, and Arabic. Arabic has more syllabus per digit than all of the other languages examined.
 * Task 1: reading speed - measures how long it takes to read words in each language. Predict that digit span will follow this pattern
 * [[image:reading speed.png width="353" height="205"]]
 * Task 2: digit span
 * Digit span is lowest in Arabic. Suggest that digit span is lower for Arabic-speaking participants because the letters take longer to pronounce/store in articulatory rehearsal loop
 * [[image:digit span.png width="357" height="199"]]
 * IQ scores may vary across cultures even if intelligence doesn't
 * Because the articulatory rehearsal loop is based on word sounds, span tasks are easier when number words are pronounced faster
 * Visuospatial Buffer
 * Evidence that articulatory rehearsal loop and visuospatial buffer are separate stores:
 * Visuospatial working memory ability...
 * is not impaired by making concurrent verbal responses
 * is impaired by making concurrent visuospatial responses.
 * And vice versa
 * Baddeley's model has been very successful
 * Separate storage of visuospatial and acoustic information
 * An executive at the head of the system
 * Performs memory and attention functions as well as simple storage

Christopher and MacDonald (2005):
 * Capacity of working memory**
 * George Miller's "magic number"
 * short-term store: 7 ± 2 items///chunks//
 * Measuring working memory:
 * Digit-span task: simple rehearsal with the slave systems (ex: repeat numbers back)
 * Miller's magic number: 7 ± 2
 * Operation-span task: measure the capacity of working memory when slave systems are busy
 * Example: do a math problem before hearing each word, then repeat word list back at the end)
 * About 3-4 items
 * Intelligence**
 * Working memory is the most task-general resource we have. Relevant any time we think, learn, or remember
 * More capacity makes us better thinkers
 * Working memory performance is related to:
 * Overall intelligence
 * Verbal fluency, reasoning ability, reading ability
 * grades in school, income
 * The important thing is the ability to //use// working memory
 * These are correlated with tasks like //operation span//, not digit span
 * Major depression is related to:
 * difficulty concentrating
 * unwanted negative thoughts
 * Two groups of participants: Depressed and matched controls (hospital employees)
 * Completed a battery of WM tasks (articulatory, visuospatial, central executive)
 * Control group performed better than depressed group across the board.

Three stages of long-term memory Rehearsal Levels of Processing
 * Acquisition of LTM & Depth**
 * Acquisition/Encoding
 * Storage
 * Retrieval
 * Glenberg, Smith, & Green (1977): Participants studied a 4-digit number for 2 seconds. Then they rehearsed a word for 2s, 6s, or 18s. Given a surprise test for the word
 * Result: Rehearsal doesn't help when you're doing shallow processing
 * Shallow processing
 * Thinking about superficial characteristics (e.g., typeface, letters, sound)
 * Deep processing
 * Thinking about meaning
 * Craik and Tulving (1975): Subjects study 60 words
 * 3 types of tasks:
 * Is the word lowercase?
 * Is the word a rhyme?
 * Does the word fit a category?
 * Result: Deeper processing, better performance
 * [[image:rehearsal.png width="267" height="160"]]
 * The ultimate level of processing:
 * Nairne et al. (2008): Participants rate words on pleasantness, imagery, self-reference...and relevance to survival
 * Survival condition performed best
 * Encoding depends on how much time you spend rehearsing in a //meaningful// //way//

Laptops Mueller & Oppenheimer (2014):
 * Transcription versus understanding**
 * Experiment 1: Participants watched a 15 minute TED talk
 * Group 1: Took notes with laptop (not allowed to multitask)
 * Group 2: Took notes with pen and paper
 * Thirty minutes later, they took a test (without studying their notes)
 * Factual questions (ex: About how many years ago did the Indus civilization exist?)
 * Conceptual questions (ex: how do Japan and Sweden differ in their approaches to equality within their societies?)
 * Results: Factual performance wasn't affected. For performance on conceptual questions, laptop users performed worse than the longhand group. Laptops hurt even if you don't multitask
 * Laptop users take more notes, and have more verbatim notes
 * Laptop means you...
 * Take a lot of verbatim notes
 * almost transcribe
 * Longhand means you...
 * Think
 * Elaborate
 * Make connections
 * Impose your own organization
 * Experiment 2: warned participants of why laptop users tend to do worse
 * Result: warning participants not to transcribe didn't solve the problem
 * Experiment 3: Participants got to study their notes this time
 * Result: laptop users still performed worse