Mannheim Master in Data Science
Spring Term 2024
University of Mannheim
Welcome to this course!
M
to open the menu?
on your keyboard to learn more about how to navigate the slides!Resources for this session
Episode 1 of CCS Podcast (Domahidi & Haim, 2021); Big Data, New Epistemologies and Paradigm Shifts (Kitchin, 2014)
What are the main questions to be answered in this session?
Course Website
Course material
About Felix
My research interests include:
What about you?
Please be open to…
Let’s solve some interesting problems together!
Expertise
Methodological competence
Personal competence
Live Sessions
Reading Task
How to read, and how not to read?
Exercises (not graded)
Graded Performance (“Prüfungsleistung”)
Structure of Research Report
Live Sessions
Date | Topic |
---|---|
Apr. 12 | Introduction Opportunities, Challenges, and Pitfalls of Computational Communication Science How to Develop a Research Problem in Computational Communication Science? |
May 3 | Project Weekend |
May 4 | Project Weekend |
May 5 | Project Weekend |
May 17 | Project Presentation |
Any Questions?
https://www.gocomics.com/calvinandhobbes/1990/06/09
Resources for this session
When Communication Meets Computation: Opportunities, Challenges, and Pitfalls in Computational Communication Science (van Atteveldt & Peng, 2018)
According to van Atteveldt & Peng (2018), CCS studies generally involve:
Three components according to Geise & Waldherr (2021):
So basically: New data & new methods, but same media & communication?
The Lasswell Formula: “Who says what in which channel to whom with what effect?”
Question | Element | Analysis | Example | CCS |
---|---|---|---|---|
Who? | Communicator | Control Analysis | Dory | ? |
Says What? | Message | Content Analysis | “Just keep swimming!” | ? |
In Which Channel? | Medium | Media Analysis | F2F (?) | ? |
To Whom? | Audience | Audience Analysis | Marlin (?) | ? |
With What Effect? | Effect | Effects Analysis | ? | ? |
Communication as the transmission of information…
…or the creation of meaning? (Berger & Luckmann, 1966)
“[Big data analytics] threatens to colonize the social sciences and humanities by turning these fields into computer science. If computational methods enter the curriculum of communication studies degrees in a major way that requires students to learn advanced programming, then not enough time will be left for practicing critical thinking, qualitative methods, social theory, critical theory, ethics, philosophy, history, and other crucial liberal arts skills because learning how to code properly is very time-intensive.”
— Fuchs & Qiu (2018)
So, what about theory?
According to the APA Dictionary of Psychology…
Is theory opposed to method then?
“There is nothing so practical as a good theory”
— Lewin (1951)
“There is nothing so theoretical as a good method”
— Greenwald (2012)
Analysis of two decades of Nobel awards in physics, chemistry, and medicine revealed that…
The data deluge makes the scientific method obsolete (Anderson, 2008)
“We can stop looking for models. We can analyze the data without hypotheses about what it might show. We can throw the numbers into the biggest computing clusters the world has ever seen and let statistical algorithms find patterns where science cannot.”
— Anderson (2008)
Four paradigms of science (Kitchin, 2014)
Paradigm | Nature | Form | When |
---|---|---|---|
First | Experimental science | Empiricism; describing natural phenomena | pre-Renaissance |
Second | Theoretical science | Modelling and generalization | pre-computers |
Third | Computational science | Simulation of complex phenomena | pre-Big Data |
Fourth | Exploratory science | Data-intensive; statistical exploration and data mining | Now |
Experimental science = pre-renaissance? 🧐
“Kuhn’s proposition was that a paradigm constitutes an accepted way of interrogating the world and synthetizing ideas and knowledge common to a substantial proportion of researchers in a discipline at any one moment in time”
— Kitchin (2014)
Thomas Kuhn (physicist & philosopher, “The structure of scientific revolutions”)
“Because of their high level of abstraction, macro theories are often criticized for not being easily amenable to empirical research. For example, it is hard to directly derive testable hypotheses from abstract, overarching sociological concepts”
— Waldherr et al. (2021)
But, they can be helpful to answer questions such as:
Answering these questions can be very helpful when evaluating or planning CCS research!
based on van Atteveldt & Peng (2018)
Relationship between comparative advantages (rows) and recommendations (columns)
Provide Multi-Causal Inventories | Report Informative Findings | Measure Important Variables “As Is” | Use Purposive Samples | |
---|---|---|---|---|
Test External Validity | Fully articulate state of prior knowledge | Give “null finding” equal importance | Converts construct validity concerns to evaluations of theoretical boundary conditions | Guides field to deploy systematic, controlled variation across studies |
Explore Theoretical Relevance | Identify when prior expectations are weak for relevant variables | Identifies theories that do and do not explain real world phenomena | ||
Audition Hypotheses for Field | Provide all hypotheses that fit the data | Show that an effect holds in a specific, defined, sub-group and so may be generalizable | ||
Create Unimaginable Hypotheses | Researchers must go beyond “most intuitive” explanation and acknowledge competing alternatives | Encourages theoretical interrogation of unexpected predictors | Reduce reliance on face valid comparisons that disguise counterintuitive differences |
based on Jungherr & Theocharis (2017)
Any Questions?
https://www.gocomics.com/calvinandhobbes/1993/01/08
Resources for this session
Forschungsmethoden und Evaluation in den Sozial- und Humanwissenschaften (Döring, 2022)
Get inspiration from publications of scholars who are active in the (C)CS community or from divisions and interest groups of the ICA.
Check out Communication Science Journals (maybe search for terms related to CCS)
Try to answer these questions:
Try to answer these questions:
Any Questions?
“Optimal psychological functioning and experience”
Ryan & Deci (2001, p. 142)
in the context of computer-mediated communication, see Meier & Reinecke (2021)
Parasocial Phenomena
From PSI to PSR
Former Student Project
We assumed that…
H1: …the intimacy of a PSR increases over a) the duration of the relationship, as well as through more frequent information disclosure b) of the persona and c) of the recipient.
H2: …the intensity of a PSR increases over a) the duration of the relationship, as well as through more frequent information disclosure b) of the persona and c) of the recipient.
Discussion
Resources for this session
The Content Analysis Guidebook (Neuendorf, 2017)
“Content analysis is a research technique for the objective, systematic, and quantitative description of the manifest content of communication.”
Berelson (1952, p. 18)
Is there a problem with this tweet?
Manifest vs. latent content
“Content analysis is a research technique for making replicable and valid inferences from texts (or other meaningful matter) to the contexts of their use.”
Krippendorff (2018, p. 18)
“Content analysis is a summarizing, quantitative analysis of messages that follows the standards of the scientific method (including attention to objectivity–intersubjectivity, a priori design, reliability, validity, generalizability, replicability, and hypothesis testing based on theory) and is not limited as to the types of variables that may be measured or the context in which the messages are created or presented.”
Neuendorf (2017, p. 17)
Can we only analyze whats “in” the text? (like the content of a container)
“The term ‘content’ in content analysis is something of a misnomer because verbal materials may be examined for content, for form (e.g., style, structure), function, or sequence of communications”
Smith (2000, p. 314)
Research objects
Potentials of content analysis
Which potential units of analysis do you see here?
Article level, paragraph level, sentence level, picture level, word level
Materials for manual coding
Example for a codebook here: https://osf.io/2z3dk/
“Content analysis is a summarizing, quantitative analysis of messages that follows the standards of the scientific method (including attention to objectivity–intersubjectivity, a priori design, reliability, validity, generalizability, replicability, and hypothesis testing based on theory) and is not limited as to the types of variables that may be measured or the context in which the messages are created or presented.”
Neuendorf (2017, p. 17)
\(measured score = true score + error score\)
Measure of intercoder reliability
\(alpha = 1 - \frac{D_{O}}{D_{E}}\)
\(D_{O}\) = observed disagreement
\(D_{E}\) = expected disagreement
For discussion of different reliability indeces, cf. Neuendorf (2017)
\(alpha = 1 - \frac{nm - 1}{m-1} (\frac{\sum pfu}{\sum pmt})\)
\(pfu\) = product of any frequencies for a given case that are different (i.e., show disagreement)
\(pmt\) = each product of total marginals
\(n\) = number of cases coded in common by coders
\(m\) = number of coders
\(alpha = 1 - \frac{(10)(2) - 1}{2-1} (\frac{3}{132})\)
\(= 1 - \frac{19}{1}(\frac{3}{132})\)
\(= 1 - 19(.023)\)
\(= 1- .43\)
\(= .57\)
Interpretation of Krippendorff’s Alpha: ≥ .667 (acceptable); ≥ .8 (good)