Please enable JavaScript.

Coggle requires JavaScript to display documents.

Digital data (Differences in traditional and digital news (publications,…

- - - - advantage: large scale, extracting hidden info, sampling not necessary (census)
      - assemble, filter and interpret data
      - disadvantage: resource demanding (data, storage, software), limited to surface-level (manifest)
      - systematic processing of big data
- - - - limited to manifest content
        (ACA cannot understand language)
      - more resources than manual coding
    - - transparency is a potential strength but in practice a weakness
      - validation not always reported (is coding accurate?)
  - - - make codebook, train computer
      - advantages
        
        100% reliable
        
        large amount of data
        
        transparent (thus valid)
      - algorithm tries to replicate human coding
      - limitations
        
        if training poor, so is computer
        
        blackbox
        
        only as good as master
    - - looks for co-occuring terms, indicating meaning of text
      - relies on clustering methods to find patterns
      - inductive
      - advantages
        
        make sense of big data
        
        not limited to key terms, looks at all words that might be of interest
      - limitations
        
        no standard to compare to (validity)
        
        interpretation still subjective task
        
        leaving certain things out, or amount of terms can have large influence
    - - limitations
        
        dictionaries are handmade: time consuming
        
        not suitable for big data
        
        not suitable for latent concepts
        
        dictionaries are topic specific
        
        limited to key terms
      - visibility analysis
      - deductive