Data Collection Analysis & Interpretation

Without good data, research would be no better than blind speculation. Hence, data collection is the most fundamental aspect of research. On the other hand, even the best data remain of little value until you light them up with analysis and interpretation. Any scientific investigation would require all three.
  1. Observation

    • Both inside and outside laboratories, researchers obtain data through observation.

      One way to collect data is through observation, notes David Nachmias and Chava Nachmias in their book "Research Methods in the Social Sciences." In experiments, you observe whether a certain stimulus produces a predicted effect. In field research, you observe how one variable, such as military rank, might relate to another, such as esprit de corps.

    Survey

    • The only way to obtain certain data is to question people.

      You can collect data by asking people questions. This can be done through surveys, states the book "Survey Research Methods," by Floyd J. Fowler. Surveys involve personal interviews, telephone interviews or mailed questionnaires, notes Fowler. The ability to generalize from survey findings depends on getting a high percentage of your subjects to respond.

    Qualitative Research

    • You can collect data through immersing yourself, either secretly or as an openly identified investigator, in the lives of the people you are studying. Such immersion may enable an inside understanding of how your subjects see themselves and their world, note Nachmias and Nachmias.

    Secondary Data

    • Sometimes data sufficient for an analysis already exist---in archives or libraries.

      In observation, surveys and qualitative research, you collect your own data. Sometimes, however, you may prefer to use data that others have already collected. These data might come from such sources as public opinion polls, governmental documents, diaries and letters.

    Analysis

    • Univariate analysis reveals how the cases stack up on one dimension.

      "Data analysis" refers to uncovering relationships among the data or, as in univariate analysis, examining data one variable at a time, according to Babbie. For example, a univariate analysis regarding first-time pregnant women might list their ages. You could make such data more manageable by summarizing it, reporting not everyone's age but only the range of ages, average age and median age.

    Multivariate

    • A graph can show clearly and simply one variable's impact on another.

      Multivariate analysis aims to show how one or more factors may affect the phenomenon of interest. For example, not just her age but also whether or not she is married and how much education she has might affect a woman's likelihood of getting pregnant. Through the statistical technique of multiple regression, you can obtain numerical estimates of how much each of two or more variables affects the outcome, suggests the book "Introduction to Probability and Statistics," by William Mendenhall.

    Interpretation

    • Analysis reveals relationships. Interpretation requires pondering why they exist.

      Reviewing the data analysis, interpretation clarifies how relationships between variables might work, according to Babbie. Your analysis might show a relationship between longevity and having Japanese parents. Interpreting the findings might include a consideration of whether genetic inheritance surpasses the Japanese diet as contributing to a long life. Interpretation might look at what other researchers have found and suggest questions for future research.

Learnify Hub © www.0685.com All Rights Reserved