Lab 5
Word Superiority and Lexical Decision
Word Superiority and Lexical Decision
Stimulus Type Proportion Correct Reaction Time (ms)
Word 0.958 1081.48
Nonword 0.958 1101.48
● How does the pattern of your individual data relate to the pattern of results predicted?
As the experiment predicted, my response time for the word trials was faster than for nonword trials. However, my accuracy for each was the same, which differs from the expectation of word trials having higher accuracy.
● What does this lab tell us about reading as a bottom-up or top-down process? What do you think would happen when learning a new language?
This experiment is a good example of how we use both bottom-up and top-down processing concurrently while reading. During this task, we use bottom-up processing as we translate the visual image, or letters, into their sounds and form the word in our mind. Our top-down processing plays a role in our recognition, enabling faster RT and better accuracy, as the word frequency theory would explain. The more familiar we are with a word from our lexicon, based on our past experiences, the quicker we are able to read, recognize, and process it (Goldstein, 2019). When learning a new language, we would not have our top-down processing to rely on for this effect. There would be no prior knowledge to connect with or make associations, and no past experience to build context or make inferences and predictions from. Lacking the benefits enabled by top-down processing, it would be challenging and time-consuming to build a new lexicon and semantics.
Condition Reaction Time (ms)
Associated 581.625
Unassociated 625.125
Nonword 674.125
● How does the pattern of your individual data relate to the pattern of results predicted?
My pattern of results is exactly as predicted, with associated words being the fastest RT, followed by unassociated, and nonwords being having the longest RT.
● What does this lab suggest in the role of top-down processing when reading? What does it reveal about a “web of concepts”?
The lab highlights the relevance of top-down processing in reading tasks, showing that prior knowledge and experience play a role in how quickly we recognize and process input based on our mental lexicon (Goldstein, 2019). Our web of concepts is arranged around this knowledge and experience, holding numerous schemas that organize related information, linking a concept with numerous properties (Goldstein, 2019). The properties within a semantic network are stored in close proximity, enabling our brain to make associations and predictions more rapidly, explaining this lab’s predictions.
● These labs are all about recognizing elements of your language quickly and accurately. We tend to think of reading as a passive skill because it’s automatized. What do you think these labs tell us about what needs to happen for artificial intelligence (think Alexa and Siri) to recognize and process language as quickly and efficiently as we do? What do you think we are still better at doing compared to AI when it comes to language?
Despite feeling like a passive skill, reading requires constant processing, ideally including top-down. Taking meaning from the text, reflecting, and thinking about it makes this an active process. AI has advanced in the language and communication arena, but still cannot perform this skill as efficiently as a human can. Our top-down allows us to make casual, instrument, or anaphoric inferences, resolve ambiguity, consider context, and make predictions based on our knowledge, past experiences, and scripts, all things AI cannot achieve (Goldstein, 2019).
AI is capable of having a far larger lexicon than a human, as well as being able to ‘learn’ new data much faster than any human, but I still think humans have the upper hand. We have the advantage of being able to follow a complicated plot and incorporate prior knowledge, and most importantly, humans have the ability to not only understand emotion being conveyed but can also have an emotional response. For these reasons, I think humans will always remain superior to AI, unless AI develops emotional intelligence and top-down processing, which is probably in the works somewhere. Scary.
References –
`Goldstein, B. E. (2019). Cognitive Psychology: Connecting Mind, Research, and Everyday Experience (5th ed.). Cengage.
MindTap - Cengage Learning. (2014). Ng.cengage.com; Cengage Learning. https://ng.cengage.com/static/nb/ui/evo/index.html?deploymentId=5868562500252548489124156979&eISBN=9781337408301&id=2075336089&snapshotId=3952969&