Pool of possible examinations questions

Pool of possible examinations questions for the written examination in PSYE60

This document includes a pool of examination questions for PSYE60 – Psychology and AI. Four of the six questions on the examinations will be taken from this pool. You are welcome to study these questions prior to the examinations. Two questions will be not be taken from these pool, so you may also need to read up on the other aspect of the course. 

The examination will be composed of 6 questions, each question is worth 4 points, which gives a total of 24 points. You need 12 for an E, 14 for a D, 16 for a C 18 for a B and 20 for an A.

 

Bring your computer to the examination

At the examination, bring your own computer, as the examination is conducted in an electronic format. If you are unable to do se, please email the Sverker Sikström prior to the examination. 

 

Social AI

One of the most important, and well-studied aspects in Social Robotics is anthropomorphism. What are its positive and negative effects on the building up of human-robot interactions, e.g., for establishing an emotional bonding.

The use of robots as social agents has enabled a new field of research for studying interactions between robots and humans (HRI). In the HRI field, theories and scientific evidence on how humans interact with and understand each other, have informed the design and implementation of interactional models between robots and humans. Describe the interactional aspects most studied and implemented by the HRI field, e.g., at the behavioural, verbal, and non-verbal level, and their impact on how humans perceive, understand, evaluate and set their expectations in interaction with robots.

 

Ethics in AI

Describe what type of regulations have been implemented/are in the progress of being implemented to ensure that ethics aspects are applied and considered in the context of AI.

 

MVPA & Cognitive Neuroscience

What is Multivariate Pattern Analysis (MVPA) and what is the general procedure for applying MVPA to brain data?

What is temporal generalization? Give examples of how temporal generalization can be used to study brain and cognitive function.

Imagine that you want to investigate how long does the brain take to retrieve a memory. How could you use multivariate pattern analysis to answer this question?

What is Multivariate Pattern Analysis (MVPA) in the context of cognitive neuroscience, and how does it differ from the traditional univariate analysis?

Explain the process of how MVPA decodes mental states from neural information. What steps are typically involved?

Consider a study in cognitive neuroscience aiming to differentiate the neural activation patterns associated with two cognitive processes. How could researchers employ MVPA to achieve this objective? Illustrate your answer with an example and hypothetical findings.

Illustrate with a specific example how an encoding model could be used to predict neural activity based on stimulus information and, conversely, how a decoding model could be used to infer stimulus or cognitive state information from observed neural activity.

Define 'encoding models' and 'decoding models' and explain how such models can be used in brain-computer interfaces.

 

Artificial Intelligence for Mental Health and Mental Illnesses

Why might an AI-driven therapist be more effective than a human psychotherapist?
Name a few of the ways that AI can support healthcare
What are some of the benefits and disadvantages of therapeutic chatbots?

 

Face recognition and AI

Discuss what you believe are the most important ethical issues involved in studies relying on facial pictures.

 

AI history

AI was founded as field in the mid 1950. The promises of AI have sometimes been very positive and sometimes negative. Summarize the reason for why that is the case!

 

Abstract Neural networks

How does forgetting occur in neural network?

Describe human memory phenomena that can be accounted for by forgetting in neural networks! Also describe how it can be accounted for.

Describe in detail how retrieval of a memory occurs in an abstract neural network.

 

Deep Neural Networks

Why do we need deep neural networks? Give an example where we need, and one where we do not need, a deep neural network to simulate a psychological a phenomon.

Describe learning in a deep neural network.

 

NLP models

Describe, in detail, more than one AI method that can be used to assess psychological constructs from open-ended language responses!

Mental states are commonly measured by rating scales in behavioral science. However, they can also be measured prompted languages responses that are analyzed by AI methods. Compare the advantages and disadvantages of these two methods!

Discuss psychological phenomena that can be found in large language models and phenomena that is difficult to find in them!