AN INTRODUCTION TO CONTENT ANALYSIS 9310021A SALLY 9310001A HELEN 9310003A STAN CONTENT ANALYSIS AS A TECHNIQUE - for making theory - by analyzing, examining, and selecting data - systematically & objectively CRITERIA OF SELECTION - clearly & fully expressed rules - set up before analysis - explain various data completely - applied strictly CATEGORIES = MAJOR POINTS = PROS & CONS SHOULD BE - connected with what is being discussed in the messages - exact wording used in the statement SHOULD NOT BE - based on personal opinions - irrelevant to the messages QUANTITATIVE 量化 vs. QUALITATIVE 質性 - Quantitative : objective, systematic, procedures of analysis arbitrary limitation, relevant categories - Qualitative : definitions, symbols, detailed explanations, etc no absolute truth, but context-bound MANIFEST vs. LATENT CONTENT ANALYSIS - manifest content (surface structure): perceptible, clear, comprehensible message - latent content (deep structure): implied, unstated message COMMUNICATION COMPONENTS 1. message ↓ 2. Sender (participants) ↓ 3. Audience (interviews) - in vivo codes: wording that participants use in interview - constructed codes: coded data from in vivo codes, created by researcher, academic terms LEVELS & UNITS OF ANALYSIS words, phrases, sentences, paragraphs, sections, chapters, books, ideological stance, subject topic, elements relevant to the context SAMPLING STRATEGIES I. Random Sampling 1. Simple Random Sampling 簡單隨機抽樣 to draw subjects from an identified population (母群體) 2. Systematic Sampling 系統抽樣 (Interval Random Sampling 間隔隨機抽樣) select nth name from the population Population 母群體總數 Sampling interval = Numbers of persons desired 抽樣間隔 取樣數目 * Random Numbers Table 亂數表 3. Stratified Sampling 分層抽樣 - divide population into stratum - ensure : dissimilarity between stratum ↑ similarity inside of each strata ↑ ∴ produce a representative sample II. Non-random Sampling Purposive Sampling 立意抽樣 researcher select subjects according to his/her research purpose and understanding of the population - researcher: with sufficient knowledge or expertise - subjects: represent the population GROUNDED THEORY 紮根理論 a process of constructing: various data →induction/deduction→theory * explain the phenomena - development of theory - collect, analyze, & compare data systematically - theory is grounded on data 7 MAJOR ELEMENTS IN WRITTEN MESSAGES 1. Words – the smallest unit, frequently used 2. Themes: simple sentences, string of words with S + predicate (e.g. You are beautiful) 3. Characters: persons 4. Paragraphs: difficult to classify ∵ various things are stated & implied in a single paragraph, infrequently used 5. Items: books, letter, diary, etc 6. Concepts: an idea, more latent e.g. crime 7. Semantics: how affected the words may be Combinations of Elements Interview #1 Ah…I do not think I improve grammar and word dictions because my teacher did not correct my grammar and word dictions. Actually, I know I am not good at writing, and I really want to improve my writing ability. Hmm……However, I also wrote articles which were asked from professors as homework while I wrote dialogue journal writing. Well, for the first time, I can accept that I had so many writing mistakes, and I know I still have room to improve it after teacher’s correction. Unfortunately, after many times corrections, the articles which were corrected by professors still appeared many grammar problems and sometimes had word dictions problems. This is why I do not think dialogue journal writing can improve our writing ability. (Shake head) Units and Categories Units = Codes ‘Code’ the elements into ‘Inductive Categories’ ex. Words, items, themes… Classes and Categories 3 major procedures: 1. Common classes 2. Special classes 3. Theoretical classes Classes and Categories Common Classes: -- a culture in general People in society to tell apart persons, things, and events Ex. Age, gender, mother… Classes and Categories Special Classes: -- the labels used by members of certain areas to tell apart persons, things, and events within their limited province out-group – people in society in-group – people in the specific group Classes and Categories Theoretical Classes: -- emerge in the process of analyzing the data -- Function: 1. grounded in the data 2. Get a theory Open Coding 1. Major Problems: -- can not read between the line -- do not get the real motivation 2. Can get the points continue. Coding can Open Coding 4 basic guidelines: 1. --- Ask the data a specific and consistent set of questions. What study are these data suitable? What category does this incident indicate? Benefits: -- sometimes find unexpected results Open Coding 2. Analyze the data minutely. categories, incidents, interactions, and the like be coded <during open coding> extensive theoretical coverage <be thoroughly grounded> systematic coding Open Coding 3. Frequently interrupt the coding to write a theoretical note. -- comments ideas <take notes> 4. Never assume the analytic relevance of any traditional variable until the data show it to be relevant. -- any traditional variable ex. Age, sex, social class… -- earn their way into the grounded theory Coding Frames Purposes: 1. To organize the data after open coding has been completed 2. To identify findings Coding Frames Axial Coding: 1. Different ideas construction 2. New ideas organize and Coding Frames Data MJ 1 Open coding Axial Coding MJ 2 MJ 1 Axial *MJ=Major Point MJ 3 MJ 3 A Few More Words on Analytic Induction Involve several refinements. Glaser and Strauss suggest: - Combine 2 data analysis. 1. Analysis of data after coding. 2. Analysis of data while integrating. Incorporate all appropriate modes of inquiry: Induction, deduction, and verification Interrogative Hypothesis Testing 4 steps of negative case testing: 1. Make a rough hypothesis. 2. Conduct a thorough search. 3. Discard or reformulate hypothesis. 4. Examine all relevant cases. 4 Safeguards against the potential flaws 4. Examples should be lifted at random. Assertion should be more than 3 examples. Analytic interpretations should be examined by independent reader. Check no invalidated overall patterns. Use safeguards can avoids “exampling” 1. 2. 3. STRENGTHS AND WEAKNESSES OF THE CONTENT ANALYSIS PROCESS Advantages: 1. It can be virtually unobtrusive. 2. It is cost effective. 3. It provides a means of study a process. Weaknesses: 1. Limited to examining already recorded messages. 2. Ineffective for testing causal relationships between variables. 3. Not appropriate in every research situation. COMPUTERS AND QUALITATIVE ANALYSIS Using qualitative research programs: 1. Help qualitative sorting and data management. 2. Takes times to learn. 3. Researchers still need to think. 4. Offer clear directions for novice. Quantitative research programs help researchers to deal with the vast number of statistical data.