Currently I am working on the portrayal of immigration and immigrants in US newspapers from the 1870s to the 1920s. As my prime source I use ProQuest Historical Newspapers. For each decades the search (I use keywords: immigration or immigrant) produces several thousands of texts. As I'm most interested in the discourse, using the method of discourse analysis and frame analysis. Therefore I concentrate on editorial articles, however, a lot of interesting information is also in news pieces. Therefore I wonder what sample size I should use to reduce the results to the size that I could work on. My research is qualitative, however, I want to make sure that the content analysis meets its requirements.
Qualitative analysis does not use sampling methodology, as such, because it is not the goal to generalize. But as you point out, the data analysis must be manageable.
The concept you need to explore is "data saturation." In other words, do you have enough individual texts to provide a clear picture of your subject matter?
In my doctoral dissertation, I also did a content analysis of news stories (see attached), but my topic was more narrow and I ended up with a total of 84 stories for analysis and did not need to sample.
First I would consider whether you should delimit your study further. I can imagine that there may have been great changes in how immigrants were portrayed in 1870 compared to 1920. Do you need the entire 50-year range? Do you need the entire geography, which was much greater in 1920 than it was in 1870?
Then figure out how to demonstrate data saturation. Maybe select a sample of 100 and perform your analysis...and then select an additional 20, and see if they change the findings significantly (or whatever numbers fit your resources).
Thesis Portrayal of The Mission Of Higher Education In The Media: A...
Thank you for some really good advice! You are right that I chose to research on a very long period of time. I have decided to divide it into smaller case studies focusing on the major events concerning immigration to the US (e.g. the passing of immigration acts, world war I, the 'red scare" etc.). This helped me to divide the gathered material (ProQuest Historical Newspapers produced thousands articles for each case study). With each I will do as you advise, that is I will try to use data saturation. Once again, thank you!
I adopted qualitative content analysis during my doctorate dissertation (copy of dissertation available on Researchgate). Please note that selecting sample size in qualitative content analysis research should focus in uncovering phenomena in context and situation. It is not aimed at generalization. Findings are limited to the data consulted. Consequently, if you have a large data you may wish to consider trimming your research questions to only a few. In the alternative, consider reducing the geographical location, or the period. Best wishes as you progress in your doctorate studies!
I also did my PhD with qualitative content analysis and social imaginary perspective. I collected more than 500 piece of news from the 3 most important newspapers in order to the Rankin of selling. In this case I have chosen 6 last years.
I think you should determined the time, kind of paper, etc. And when the election is completely justified, to analyse all of them into this period of time.
Good luck!
Thank you all for your help! Do you know any good handbook on qualitative content analysis of media?
QDA is usually based on an interpretative philosophy. QDA is the range of processes and procedures whereby we move from the qualitative data that have been collected into some form of explanation,understanding or interpretation of the people and situations we are investigating. The idea is to examine the meaningful and symbolic content of qualitative data .The key point is this: Our qualitative sample must be big enough to assure that we are likely to hear most or all of the perceptions that might be important. If the team wants to be more conservative and reduce the risk of missing the one-in-100 subgroup then a sample of nearly 500 would be needed.
It's not a handbook, but in terms of resources, my first recommendation is Krippendorff's (2004) "Content analysis: An introduction to it's methodology."