Research Methods (8-)
Extraneous variables
Any variables, in addition to the independent variable, that affect the dependent variable. If effects are random, add unwanted error variance, decreasing ability to find effects. Age, research assistant differences, a strong predictor of the dependent variable (e.g., pre-existing attitudes, physical condition). Can "control" for this error variance if you measure these variables by partialing out the variance they contribute.
Surveys
A collection of questions given to respondents. Used in experiments & correlational studies
"p value"
In Inferential stats we figure out the likelihood that we could have gotten the differences we find in our sample just by chance if the null hypothesis is true. Can be anything from (almost) 0 to (almost) 1. E.g., If p = .4, there's a 40% chance we could have found what we did randomly if there was no real difference (i.e., you can only be 60% sure that the relationship you found is a real one)
Substantive Theory
Observed patterns are related to each other and a theory is developed to explain the connections in that setting. (Outcome of Grounded Theory)
Audio and video recording
Review verbal and non-verbal details. Involves a lot of transcription
Curvilinear regression analysis
A form of regression analysis that allows relationships among variables to be expressed with curved geometric lines instead of straight ones.
Ongoing access
Can be problematic b/c People get suspicious of the researcher's motives. Group members fear that what they say or do will get back to bosses or colleagues. People being studied may decide to sabotage the research.
Nominal
Categorically discrete data -it is for mutually exclusive, but not ordered, categories. E.g., Names of political parties, Names of different kinds of jobs, gender etc.
Structured
Design a schedule of things researchers want to observe for the study
Code notes
Indicate the code labels, the meaning and definition of the codes, and detailed information about the different types of coding
Rapport
Interviewers should try to develop rapport with interviewees but maintain a cautious balance. 'Too friendly' may has cause the interview to get side-tracked, go on too long, or bring an interviewee to tailor their responses toward 'pleasing' the interviewer.
Standard Error
Involves two essential components: the confidence level (for example, 95 percent) and the confidence interval (for example, plus or minus 2.5 percent).
Prompting
Occurs if the interviewer suggests a specific answer for the interviewee. Prompting should only be used as a last resort when a respondent absolutely cannot come up with a reply.
Daily Diaries
Questionnaires to the next level -regularly fill out set questions. Can get really good, accurate data. But people may drop out, expensive.
Naturalism
An approach to field research based on the assumption that an objective social reality exists and can be observed and reported accurately.
Grounded Research Theory
Begins as an inductive type of research, based or "grounded" in observations. Attempt to derive theory from analysis of patterns, themes, and common categories discovered in observational research (i.e., don't start project with a preconceived theory in mind). Collect data, code it, then derive theory
Central tendency
Beyond simply reporting the overall distribution of values, you may choose to present your data in the form of central tendency. Mode (good for all types!) Median (not good for nominal) Mean (not good for nominal or ordinal)
Closed
I.e., private or restricted settings: meetings of private clubs, social movement organizational centres, etc.
Interval
Is like ordinal except we can say the intervals between each value are equally split-the difference between two values is meaningful. E.g., temperature in degrees Celsius, age in years, income
Field notes
Structured or unstructured. Write down your observations as soon as possible.
integrative memos
seek to clarify and link analytic themes and categories.
Tests of statistical significance
A class of statistical computations that indicate the likelihood that the relationship observed between variables in a sample can be attributed to sampling error only. See also inferential statistics.
Frequency distribution
A description of the number of times the various attributes of a variable are observed in a sample. The report that 53 percent of a sample were men and 47 percent were women would be a simple example of a frequency distribution. Another example would be the report that 15 of the cities studied had populations under 10,000, 23 had populations between 10,000 and 25,000, and so forth.
Reactivity
A difficulty that might arise when subjects of social research react to the fact of being studied, therefore altering their behaviour from what it would have been normally.
Questionnaire
A document containing questions and other types of items designed to solicit information appropriate to analysis. Questionnaires are used primarily in survey research but also in experiments, field research, and other modes of observation.
Case study
A focused, detailed investigation of a single instance of some social phenomenon like a town, an industry, a community, an organization, or a person.
Case Study
A focused, detailed investigation of a single instance of some social phenomenon like a town, an industry, a community, an organization, or a person. It concerns what the unit the researchers will focus on --not how the data will be gathered (i.e., it is more concerned with what researchers are observing, not how they are observing...it is not a method per se). Exploration of a case over time through detailed and in-depth data collection. Particularly useful in areas of study where universal principles are thought to exist (e.g., neuroscience -Phineas Gage)
Oral History
A form of historical research that uses in-depth interviewing to gather data. Based on individual's recollections of the past -typically focusing on specific events or periods of time. The use of oral histories allows the researcher to examine issues in depth, which makes the method strong on validity. The issue of reliability in oral history is complex and often depends on the topic under study and the interpretations of data being offered.
Partial regression analysis
A form of regression analysis in which the effects of one or more variables are held constant (controlled), similar to the logic of the elaboration model.
Secondary analysis
A form of research in which the data collected and processed by one researcher are reanalyzed—often for a different purpose—by another. This is especially appropriate in the case of survey data. Data archives are repositories or libraries for the storage and distribution of data for secondary analysis.
Linear regression analysis
A form of statistical analysis that seeks the explanation for the straight line that best describes the relationship between two ratio variables.
Interview schedule
A formal list of questions that the interviewer must follow in detail. The questions must be asked in the order given.
Contingency table
A format for presenting the relationships among variables as percentage distributions. See 14 for several illustrations and for guides to making such tables.
Coding scheme
A framework that lays out key concepts, their definitions, and the criteria one uses for identifying the concept. This scheme or framework evolves over time in qualitative research, during the process of coding and analyzing the data. As the researcher proceeds with successive passes through the data, the scheme is subject to refinement and change.
Statistical significance
A general term referring to the likelihood that relationships observed in a sample could be attributed to sampling error alone. See also tests of statistical significance.
Coding (2)
A key procedure used by researchers in organizing and processing qualitative data, coding refers to applying labels to strips of data that illustrate ideas and concepts and to the continuing process of identifying, modifying, and refining concepts and categories that sustain emerging themes and patterns.
Standard deviation
A measure of dispersion around the mean, calculated so that approximately 68 percent of the cases will lie within plus or minus one standard deviation from the mean, 95 percent will lie within plus or minus two standard deviations, and 99.9 percent will lie within three standard deviations. Thus, for example, if the mean age in a group is 30 and the standard deviation is 10, then 68 percent have ages between 20 and 40. The smaller the standard deviation, the more tightly the values are clustered around the mean; if the standard deviation is high, the values are widely spread out.
Oral history
A method that uses in-depth interviews as a means of gathering data about the past from individuals' recollections, typically focusing on specific events or periods of time.
Probes
A nondirective phrase or question used to encourage a respondent to elaborate on an answer (e.g., "anything more?" and "How is that?"). Sometimes probes are used if the respondents need help with their answers. If used, probes should follow a standardized format. Often repeating the question or answer choices will suffice. (Psychology example: Probing for suspicion "Did you notice anything unusual?" if they answer no, ask: "Do you have any suspicions?")
Respondent
A person who provides data for analysis by responding to a survey questionnaire.
Theoretical saturation
A point in time when nothing new is being learned
In-depth interview study
A research design where qualitative interviewing is the primary means of data gathering.
Extended case method
A research technique developed by Michael Burawoy that uses case study observations to discover flaws in, and then modify, existing social theories. This technique emphasizes rebuilding or improving theory rather than approving or rejecting it.
Institutional ethnography
A research technique that uses the personal experiences of individuals (especially women and other members of subordinated groups) to uncover the institutional power relations that structure and govern their experiences.
Range
A simple example of a measure of dispersion, composed of the highest and lowest values of a variable in some set of observations. E.g., So we may report that the mean age of a group is 37.9, and the range is from 12 to 89.
Multiple regression analysis
A statistical analysis that provides a means of analyzing the simultaneous impact of two or more independent variables on a single dependent variable. The analysis produces an equation that represents the several effects of the multiple independent variables on the dependent variable.
Contingency question
A survey question intended for only some respondents, determined by their responses to some other question. For example, all respondents might be asked whether they belong to the Cosa Nostra, and only those who said yes would be asked how often they go to company meetings and picnics. The latter would be a contingency question.
Probe
A technique employed in interviewing to solicit a more complete answer to a question. It is a nondirective phrase or question used to encourage a respondent to elaborate on an answer. Examples include "Anything more?" and "How is that?"
Confounding variables
A type of extraneous variable that exerts a systematic influence on the dependent variable and therefore seriously affect the internal validity of a study. Affects the dependent variable in a systematic way that is inseparable from the influences of the independent variable.
Continuous variable
A variable where small increments in the values of the variable are logically possible, such as income. A variable whose attributes form a steady progression, such as age or income. Thus the ages of a group of people might include 21, 22, 23, 24, and so forth, and could even be broken down into fractions of years.
Discrete Variable
A variable whose attributes (values) are separate from one another. Another term is "categorical." Examples: Gender and religious affiliation. Only modes can be calculated for nominal data. Only modes and medians can be calculated for ordinal data. Modes, medians, and means can be calculated for interval or ratio data.
Discrete variable
A variable whose attributes are separate from one another, or discontinuous, as in the case of gender or religious affiliation. In other words, there is no progression from male to female in the case of gender.
Average
An ambiguous term generally suggesting typical or normal—a central tendency. The mean, median, and mode are specific examples of mathematical averages referred to as measures of central tendency.
Constant comparative method
An aspect of the grounded theory method in which observations are compared with one another and with the evolving inductive theory.
Mean
An average computed by summing the values of several observations and dividing by the number of observations. If you now have a GPA of 4.0 based on 10 courses, and you get an F in this course, your new grade point (mean) average will be 3.6.
Mode
An average representing the most frequently observed value or attribute. If a sample contains 1,000 Protestants, 275 Catholics, and 33 Jews, Protestant is the modal category.
Median
An average representing the value of the "middle" case in a rank-ordered set of observations. If the ages of five men were 16, 17, 20, 54, and 88, the median would be 20. (The mean would be 39.)
Concept
An idea or a general abstraction of one aspect of social life (example: "social action" or "social structure")
Grounded theory method (GTM)
An inductive approach to research introduced by Barney Glaser and Anselm Strauss in which theories are generated from an examination of data, through the constant comparing of unfolding observations.
Grounded theory
An inductive approach to social research that attempts to derive theory from an analysis of the patterns, themes, and common categories discovered in observational data. This differs from hypothesis testing, in which theory is used to generate hypotheses to be tested through observations.
Focus group
An interviewing method where a number of subjects are brought together to discuss a specific topic or issue. A focus group is typically led by a moderator who helps to facilitate discussion and ensures that no person dominates the conversation, while interfering as little as possible in the discussion.
Laziness or boredom
Answers just to get the end of the process. Ensure the correct answers are not all the same
Social desirability
Avoiding appearance of being bad/unlikeable. Answers are not sincere, e.g. questions on racism.
Ambiguity
Be specific (Bad: "Do you agree or disagree with the new exam regulations?" Better: "The college is proposing that the final exam period should be reduced to 10 consecutive days. Do you agree or disagree?")
Multiple-choice items
Choice among multiple alternatives (e.g., please indicate your current relationship status)
Dichotomous items
Choice among two alternatives (Yes/No) Forces an answer in one direction or another. Can be simple to analyze. But can cause restricted range of responses (Small variance in responses).
Constant comparison
Constant comparison (of data and concepts). Constant comparative method: an aspect of the grounded theory method in which observations are compared with one another and with the evolving inductive theory. Concepts develop, hypotheses generated, tested against further data (i.e., it becomes an alternation between induction and deduction -a back and forth process between data collection and data analysis).
Axial Coding
Data are reviewed for linkages and re-organized according to those connections. Specific concepts and categories refined. New codes may be developed.
Ratio
Data is interval data with a natural zero point. E.g., time and money is ratio since 0 time and 0 amount money is meaningful.
Operational notes
Deal with issues of method and procedure of gathering data and the situations encountered in the process
Measures of association
Descriptive statistics summarizing the relationships between variables.
Open
E.g., public parks, downtown sidewalks, etc. In an 'open' setting may be difficult to make observations and talk to people.
Acquiescence
E.g., respondent agrees just to be 'cooperative'. Instruments with multi-item measures are designed with items that have logically opposite positions.
Matrix items
Easier for the respondent to answer because same scale throughout. A set of scale items but all using same scale. Avoid "response set" by varying direction (i.e., use both positively and negatively worded questions) (Strongly agree, agree, no opinion, don't agree)v
Theoretical notes
Elaborate on conceptual meanings, establish relationships among concepts, and lay out theoretical propositions
Filter and Contingency items
Filter questions divide subjects into subgroups and then contingency questions are given depending on the response to the filter questions. Enables researcher to ask participants questions that are only relevant to them. (Do you smoke? Yes-->When did you start? OR No-->Do you mind if others smoke?)
Scale items
Fixed alternatives on a scale (called "Likertscale"). E.g., how much do you smoke? Two opposing ends and usually odd number of points. Sometimes even number when want to force participants to choose one side (i.e., no midpoint). Better at capturing differences between people (improved data analysis = Greater variance in responses)
Frequency distribution
Frequency, relative frequency, cumulative frequency. When the variable is discrete (nominal or ordinal) we get frequency distributions. Frequency distribution:a listing of categories or possible values for a variable, together with a tabulation of the number of observations in each category (reporting the overall distribution of values) (e.g., 45 males and 55 females participated). Can also do frequency distributions on interval or ratio variables, but less common.
Chi Square (test of statistical significance)
Helps determine if the differences in the values on the dependent variable along categories of the independent variable are significantly different from the values we would find if there was no relationship between the independent and dependent variables. The level of significance indicates the probability of obtaining a sample with a degree of association at least as great as that in the observed sample data from a population in which there is, in fact, no relationship. Used for nominal & Ordinal data.
Correlation Coefficient (r)
Helps summarize and describe characteristics of a relationship between two variables. R score provides an indication of: Direction of the relationship, + = variables move together in same direction, -= variables move together, but in opposite direction (also known as an inverse relationship). Strength of the relationship (from 0 to 1): 0 = no relationship, no pattern, 1 = perfect relationship, all points fall on straight line.
alpha-level
In Social Science, we set an arbitrary cut-off point of .05 (sometimes .01 or .001). If our p-value is less than alpha (usually .05), we say it is so unlikely to be due to chance, that it's probably a real effect.
Null hypothesis
In connection with hypothesis testing and tests of statistical significance, that hypothesis that suggests there is no relationship among the variables under study. You may conclude that the variables are related after having statistically rejected the null hypothesis.
Qualitative interview
In contrast to a survey interview, a qualitative interview allows the researcher to pursue issues in depth and gives the respondent more freedom to direct the flow of conversation. The researcher typically has a general plan of inquiry but not a standardized set of questions that must be rigidly followed. Respondents' answers to initial questions shape a researcher's subsequent questions. This method is flexible, has high face validity, produces speedy results, and is inexpensive
Level of significance
In the context of tests of statistical significance, the degree of likelihood that an observed, empirical relationship could be attributable to sampling error. A relationship is significant at the .05 level if the likelihood of its being only a function of sampling error is no greater than 5 out of 100.
Participatory action research
In this approach to social research, the researcher serves as a resource to those being studied, typically disadvantaged groups, with the aim of increasing their ability to act effectively in their own interest. Counter to the conventional status and power differences between researchers and subjects, those under study are given control to define their problems, define the remedies they desire, and take the lead in designing the research that will help them realize their aims. Collaborating with participants to change what is together regarded as a desirable direction
Self-administered surveys
In-lab (participants come to central location, receive instructions from experimenter, and complete survey on their own), Mail (surveys sent out in mail, completed by willing participants and returned via mail...self-mailing questionnaire, which requires no return envelope: when the questionnaire is folded a particular way, the return address appears on the outside. Do a follow up (resend it if they didn't send it back)) & Online (participants sign on to survey website from any computer, anywhere, and complete survey. Be careful to generalize)
proportionate reduction of error (PRE) principle
Incorporates a comparison between predicting a variable given its own distribution and given the joint distribution with another variable
Participant-observer
Integrated as a participant, but also taking observations. Often overt, but can also be covert. Presence of researcher again brings different dynamic to group or situation.
Complete participant
Integrated in the study as a full participant in the study, member of the community, etc. You let people only see you as a participant -not as a researcher. Usually covert, but can be overt (Ethical issues involved in covert researcher participation, also, presence of researcher participant brings different dynamic to group --> might alter others' behaviour). Method gets the closest to participants and their activities but there is a risk of over-identification or developing a strong dislike of the participants. Either may skew the data.
Qualitative Research
Interpretive, naturalistic approach to understanding phenomena. Studies things in their natural settings, attempting to make sense of phenomena in terms of the meanings people bring to them. Employs observation, one-to-one interviewing and documentary analysis.
Long Questions
Longer questions are confusing and harder for people to remember. They can also get bogged down with details. (Bad: Could you please tell me how many children you have, and in what year each of them was born? Better: What are the ages of your children?).
Structured Interviews
Often used because they produce standardization in the asking of questions and the recording of answers
Long Time Frames
People forget things... good to establish a time-frame (Bad: How often do you eat fast food? Better: Over the last seven days, how many times have you eaten a fast-food snack or meal?)
Wording Bias
People respond to different words: The connation of words can be different, even if they mean the same thing. (People are more in favour of policy change if you mention "poor" and "poverty". Less in favour if you mention "welfare").
Double Negatives
People tend to get confused by negatives, especially by double negatives. (Bad: Do you disagree that marijuana should be decriminalized? Better: Do you agree that marijuana should remain illegal?)
Open-ended questions
People write what they want. Advantages (Room for unusual answers, more genuine answers, Good for exploring new areas). Disadvantages (Takes a lot of time to record, code, People may not want to spend that much time). Qualitative in-depth interviewing relies heavily on open-ended questions, but they are sometimes used in other forms of data collection such as telephone surveys and self-administered questionnaires.
Historical analysis
Provides a descriptive account of a single event or historical period based on information gathered. Particularly useful for discovering patterns in the histories of different culture.
Closed-ended questions
Questions to which the respondent is asked to select an answer from among a list provided by the researcher. These are popular in survey research because they provide a greater uniformity of responses and are more easily processed than open-ended questions.
Unstructured
Record whatever elements are available in the environment that they think might be relevant
Ordinal
Refers to quantities that have a natural ordering -The values simply express an order. E.g., The ranking of favorite sports, social class (With ordinal data you cannot state with certainty whether the intervals between each value are equal)
Bi-variate Analysis
Reports findings of two variables at a time. "bi" = two, variate = variable. Look at how the variables "move" together. Generally to analyze the association between two variables. Examples include correlation, regression, t-test, ANOVA, chi-square tests.
Complete observer
Researcher separate from other participants, takes no part in community or study other than to observe. Can be overt (openly there) or covert (hidden). Ethical issues involved in covert observation. Covert has no risk of reactivity-when people react and modify their behaviour in a variety of ways because they know they are being studied. Researcher has limited information for understanding the actions of the participants.
Selective Coding
Selection of the core category / categories. Relationships among a few particular categories become the focus. Identifying gaps that need to be filled in. Conceptualize the phenomenon (emerging theory)
Content analysis
Set of methods for analyzing the symbolic content of communication. Essentially a coding operation
Informant
Someone well versed in the social phenomenon that you wish to study and who is willing to tell you what he or she knows. If you were planning participant observation among the members of a religious sect, you would do well to make friends with someone who already knows about them—possibly a member of the sect—who could give you some background information about them. Not to be confused with a respondent.
Double Barrelled Questions
Sometimes we hide two questions in one ("Do you exercise regularly to avoid heart trouble"? If someone says no, is it because they don't exercise? Or aren't avoiding heart-trouble?.) Related: Loaded questions... already assuming something! "Do your parents know you do drugs?"
Funnel items
Start broad and then get more narrow: enables researcher to obtain P's general attitudes and then more context-specific attitudes. E.g., "Some people support abortion and argue for a right to choose. Other people do not support abortion and argue for a right to life" 1.What is your opinion of abortion? 2.What is your opinion of abortion if the mother is a teenager with little external support? 3.What is your opinion of abortion if the mother will undergo serious health risk if she continues with the pregnancy?
Descriptive statistics
Statistical computations describing either the characteristics of a sample or the relationship among variables in a sample. Descriptive statistics merely summarize a set of sample observations, whereas inferential statistics move beyond the description of specific observations to make inferences about the larger population from which the sample observations were drawn. Describe either the characteristics of a sample or the relationship among variables in the sample. Simply describing/summarizing the data. Presents large sets of numbers in a manageable form. Enables researcher to explore numbers without having to read each one. Enables researcher to see patterns among numbers.
Biography
Studying an individual and their experiences as told to the researcher or found in documents (idiographic, extremely focused research)
Computer-assisted telephone interviewing (CATI)
Survey research technique in which the telephone interviewer reads the questions from the computer monitor and enters the answers directly into the computer.
Bias
That quality of a measurement device that tends to result in a misrepresentation of what is being measured in a particular direction. For example, the questionnaire item "Don't you agree that the prime minister is doing a good job?" would be biased in that it would generally encourage more favourable responses.
Generalizability
That quality of a research finding that justifies the inference that it represents something more than the specific observations on which it was based. Sometimes this involves the generalization of findings from a sample to a population. Other times, it's a matter of concepts: If you discover why people commit burglaries, can you generalize that discovery to other crimes as well?
Univariate analysis
The analysis of a single variable, for purposes of description. Frequency distributions, averages, and measures of dispersion would be examples of univariate analysis, as distinguished from bivariate and multivariate analysis. Uni = one, variate = variable.
Trivariate Analysis (Multivariate)
The analysis of the simultaneous relationships among several variables. E.g., Examining simultaneously the effects of age, gender, and social class on religiosity. Logic: Instead of one independent variable and one dependent variable, there is more than one independent variable. We seek an explanation through the use of more than one independent variable.
Multivariate analysis
The analysis of the simultaneous relationships among several variables. Examining simultaneously the effects of age, gender, and social class on religiosity would be an example of multivariate analysis.
Bivariate analysis
The analysis of two variables simultaneously, for the purpose of determining the empirical relationship between them. The construction of a simple percentage table or the computation of a simple correlation coefficient would be examples of bivariate analyses.
Typology
The classification (typically nominal) of observations with regard to their attributes on two or more variables or concepts. The classification of newspapers as liberal-urban, liberal-rural, conservative-urban, or conservative-rural would be an example.
Dispersion
The distribution of values around some central value, such as a mean, mode, or median.
Confidence level
The estimated probability that a population parameter lies within a given confidence interval. Given an appropriately constructed confidence interval, such as plus or minus two standard errors around the mean, we may state that in 95 percent of all samples the true population value will be inside the constructed interval.
Coding
The identification and labeling of concepts. Process by which classification of phenomena occurs. Labels given to issues/activities being observed that can be grouped together based on common characteristics or related meanings. Enables researchers to group similar events, objects and happenings under a common heading or classification. Starts in early stage of research project. First step in interpreting data and developing theory. Data are treated as potential indicators of concepts. The indicators are repeatedly compared for concepts / categories.
Qualitative analysis
The nonnumerical examination and interpretation of observations, for the purpose of discovering underlying meanings and patterns of relationships. This is most typical, for example, of field research and historical research.
Response rate
The number of people participating in a survey divided by the number selected in the sample, in the form of a percentage. This is also called the "completion rate" or, in self-administered surveys, the "return rate": the percentage of questionnaires sent out that are returned.
Quantitative analysis
The numerical representation and manipulation of observations (data) for the purpose of describing and explaining the phenomena that those observations reflect. "Number crunching" to describe your sample and explain relationships among your variables
Covert
The people being studied do not know they are being observed by a researcher
Overt
The people being studied know they are being observed by a researcher.
Open coding
The process of closely examining the raw data (e.g., interviews, field notes, art) in the initial stages of a qualitative data analysis with the aim of identifying, labelling, and classifying as many ideas, concepts, and themes as the researcher can without concern for how these ideas or concepts are related or how they will be used.
Open Coding
The process of closely examining the raw data (such as interviews, field notes, and art) in the initial stages of a qualitative data analysis with the aim of identifying, labeling, and classifying as many ideas, concepts, and themes as the researcher can without concern for how these ideas or concepts are related or how they will be used. Identifies initial concepts that will be categorized together later.
Memoing
The process of writing memos concerning the ideas and insights developed during the collection and analysis of qualitative data. Memos are the researcher's record of methodological concerns, descriptions and definitions of concepts, emerging or discovered relationships among codes, ideas for further study, or any other subject relevant to the study. The procedure helps the researcher to organize and process qualitative data and to discover patterns. Writing of notes and commentaries concerning ideas and patterns that emerge in the process of reading and coding data. Used to elaborate on the codes devised and summarize potential relationships that emerge among codes.
Coding (1)
The process whereby raw data are transformed into a standardized form suitable for machine processing and analysis.
Confidence interval
The range of values within which a population parameter is estimated to lie. A survey, for example, may show 40 percent of a sample favouring Candidate A (poor devil). Although the best estimate of the support existing among all voters would also be 40 percent, we would not expect it to be exactly that. We might, therefore, compute a confidence interval (such as from 35 to 45 percent) within which the actual percentage of the population probably lies. Note that we must specify a confidence level in connection with every confidence interval.
Response sets
The respondent is not motivated to provide a genuine response, : trying to please the researcher.
Inferential statistics
The statistical measures used for making inferences to a larger population from findings based on sample observations. The body of statistical computations relevant to making inferences from findings based on sample observations to some larger population. Involves hypothesis testing and making inferences from the sample to the population. Going beyond just describing/summarizing the data. Trying to draw conclusions from the sample to relationships between variables in the population. In order to determine whether you can infer that the relationships found in your sample are also there in the population (i.e., they are real), need to rule out error as the cause of the relationships. Based on probability: Null & Alternate Hypothesis. Used to estimate the expected range of error involved in inferring a population value from a sample value (Standard Error)
Conversation analysis
The study of conversation or talk-in-interaction with a focus on how talk, particularly naturally occurring talk, is managed among individuals. One key assumption of conversation analysts is that everyday talk is highly organized and ordered. Specific techniques, such as transcription conventions, have been developed by conversation analysts for the systematic and detailed study of the devices used in ordinary talk.
Semiotics
The study of signs and the meanings associated with them. This is commonly associated with content analysis.
Ethnography
The term ethnography varies somewhat in its use by researchers. It generally refers to a report on social life that focuses on detailed and accurate description rather than explanation. For some it refers also to data collected in the natural setting, while for others it refers to naturalistic observations and holistic understandings of cultures or subcultures. "Ethno" = people. Research immersed in social setting (sometimes for years). Describes a culture in detail, including language, customs, values, beliefs, religious ceremonies, and laws. Focuses on detailed and accurate description rather than an explanation. Relies on observation, participation, archival analysis, and interviewing as techniques for gathering data.
Formal Theory
Theory applied at a higher level. Requires data collection in different settings. Applicable to a variety of settings. (Outcome of Grounded Theory)
Key Informants
These are informants who are particularly knowledgeable and cooperative. Drawbacks to using them: Researcher may ignore other group members & Key informant's view may not be representative of the group as a whole.
Leading Questions
They lead people to a particular answer. (Do you think I should continue living here so we can have fun hanging out? Do you think I should move north so you can have fun visiting me when you go on vacation?)
Nonsampling error
Those imperfections of data quality that are a result of factors other than sampling error. Examples include misunderstandings of questions by respondents, erroneous recordings by interviewers and coders, and data entry errors.
Being Vague
Use shorter words and clear language for best results... even words everyone understands, like "exercise" might be interpreted differently. (Bad: How much exercise do you do each week? Better: How many minutes a week do you engage in physical activity for fitness (e.g. jogging, lifting weights, swimming)?)
Pearson's r (PRE)
Used for interval and ratio level variables and uses the mean in applying the PRE principle.
Gamma (PRE)
Used for ordinal variables and is based on comparing all possible pairs of cases for their relative rank ordering on both variables. The number of pairs of cases with the same ranking on the two variables is compared with the number of pairs of cases having the opposite ranking on the two variables. The values for gamma may vary from −1.0 to +1.0, thereby incorporating the direction as well as the strength of the association.
t-test (test of statistical significance)
Used with interval or ratio level dependent variables by examining differences in group means. The value of the t-test will be larger, showing greater likelihood of a statistically significant difference, under several conditions: when the size of the difference between the mean increases, when the size of the sample increases, and when variations of values within each group are smaller.
Lambda (PRE)
Used with nominal variables, and its calculation involves using the mode. Its value ranges from zero (no association) to one (a perfect relationship).
Extended Case Study Method
Uses case study observations to discover flaws in and to improve existing social theories. Opposite to grounded theory approach -you should start a project with a preconceived theory in mind. Goal is to improve theory rather than accepting or rejecting it
Institutional Ethnography
Uses the personal experiences of individuals (especially members of subordinate groups) to uncover the institutional power relations that structure and govern their experiences. Goal is to uncover forms of oppression that are often overlooked by more traditional types of research
Principle of Triangulation
Using multiple approaches enables an investigator to "zero in" on the "truth". Two or more different instruments or approaches provide different vantage points. Consistent findings with different methods allow more confidence about validity. Use of multiple methods cancels out problems that each method has on its own.
Intra-interviewer variability
an interviewer is not consistent in asking questions or recording answers (with the same respondent or a different one)
Causes
are the factors influencing something of interest.
Consequences
are the outcomes of a particular variable.
Initial memos
are used early in the process of analyzing data.
Curvilinear regression analysis
enables the researcher to deal with relationships that fail to meet the linear model.
Face-to-face
interviewer asks participants questions in person and participants respond verbally or on paper
Telephone
interviewer calls potential participants and administers survey via phone
Regression analysis
is a method by which the relationship between two or more variables can be specified in the form of a mathematical equation, known as the regression equation. This equation is a technique for representing the line that comes closest to the distribution of points in a joint frequency distribution. The general form of the equation in the case of one dependent and one independent variable is Y = a + bX, where X and Y represent the independent and dependent variables respectively; b represents the slope of the regression line; and a represents the Y intercept—the value of Y when X is equal to zero.
The explained variation
is the difference between the total variation and the unexplained variation.
Inter-interviewer variability
lack of consistency in asking questions or recording answers between different interviewers
Variability
range, standard deviation, interquartile range
Processes
reflect an order among the elements of structure.
Structures
reflect different aspects of a variable.
Magnitudes
reflect the levels of particular variables.
Frequencies
simply measure how often something occurs.
Reflections
take notes immediately following observations