Getting to know you… Specialization? ? Background? ? Research & Statistics experience? ? Research interest? ? HD FS 503 Research Methods in Human Development & Family Studies January 14, 2002 Susan Hegland & K.A.S. Wickrama 1 2 HD FS 503 Outcomes ? ? ? HD FS 503L Outcomes: ? Apply chain of reasoning to plan, analyze, and critique research methods Identify basic principles of sampling ? Link variables, operational definitions, measures, metrics Evaluate evidence supporting causal inferences ? ? Describe essential characteristics of qualitative, survey, and experimental methods ? ? Describe requirements for longitudinal and single subject studies Identify professional resolutions to ethical dilemmas in research ? ? ? enter, clean, describe, and summarize data using descriptive statistics and graphical displays select an appropriate statistic, test a given hypothesis, report results appropriately describe an association using correlation and regression techniques reduce data using reliability programs produce output related to normality and multicollinearity of data 3 4 Additional 503L Outcomes ? Run appropriate statistical tests in SPSS for Windows: – – – – – ? ? Overview ? ? t-test ANOVA repeated measures ANOVA MANOVA Chi-square Principles common to all research: 8 weeks Specific research methods – – – – – Report results appropriately Discuss results appropriately 5 Survey research Qualitative research Experimental research Evaluation Studies Longitudinal Studies & Action Research 6 1 Mechanics Course Questions? Syllabus? ? Assignments? ? Tests? ? Texts ? – Krathwohl – Hegland (Handbook): Copyworks – Bring to class and lab each week Web resources: find CFCS Class Web; Web CT ? – 10 terms from Krathwohl chapters – Define and give example from HD FS research – 1-2 essays: application/integration skills – http://www.public.iastate.edu/~shegland/homepage.html – Syllabus, Handbook, Adobe Acrobat lectures Office Hours: Tuesdays, 3-5 ? – Form a learning team (2 members) or I will assign you 7 8 First Learning Team Assignment: Choosing an empirical, quantitative article ? ? ? ? ? ? Necessary Background From the best research journal you know--check with a grad faculty member for ideas About a topic that you’re excited about and familiar with A study that COULD be your dissertation Using statistics and methods you’re comfortable with Make one copy for you and one for me Learning Team Assignment 2: Design, run, interpret analyses to write an article for the Journal of Research in HD FS ? ? ? ? ? ? ? Good written communication skills Good understanding of mathematics Good understanding of statistics Familiarity with research articles in human development and family studies Professional experiences in human development and family studies Basic logical analysis skills Familiarity with the APA Publication Manual 9 10 Knowledge of Research Methods ? Conceptual Knowledge – – – – Knowing “That” Understanding Networks of concepts Links between old and new information – Links to concepts at same level of abstractness – Links to more and less abstract concepts ? Building Knowledge Procedural Knowledge – Knowing “How” – Skill Concept 1 Procedure 1 – Action sequences: steps – Symbols for representing statistics – Rules for completing tasks Concept 2 Procedure 2 Concept 3 Procedure 3 – Problem solving sequences of actions 11 12 2 Building Knowledge in Research Methods Mean = (sum of y)/n Dispersion calculate S.D. Skewed Distribution calculate kurtosis Example 13 Remember what SD means (+/- 1, 2, 3 SD). Draw the following distributions with minimum score = 1 & maximum score =10 ? M = 5, SD = 1.6 ? M = 1, SD = 3 ? M = 9, SD = 3 Any ceiling effects? floor effects? What are the implications of each of these distributions? BEWARE OF MURPHY’S LAW! Aids to building knowledge ? ? ? ? ? ? 14 Link concepts ? Link abstract concepts to familiar experiences ? Link concepts to each other Automatize skills (procedures) Reflect on outcomes of procedures Look for patterns Revise concepts based on observed patterns Avoid the “plug and chug” error ? ? ? ? ? ? ? ? Computers Crash! Printers Fail! Networks Fail! Disks get Viruses! Therefore, start assignments early Work with Friends! E-mail me with problems! See me at office hours! 15 16 Introduction to Social Science Research Goals: To ? describe behavior ? explain behavior ? predict behavior Why do empirical research? 17 18 3 Exploration: ? Describing Behavior Discovers new situations Answers “what” and “how” ? Under what circumstances? ? Labeling important phenomenon ? Perceives aspects of situations missed by others ? Identifies patterns ? 19 20 Explaining Behavior Validating or predicting Follows good descriptions ? Identifying necessary and sufficient conditions ? ? ? ? ? Predicting behavior Testing our explanations Identifies – incorrect explanations – oversights – one-time occurrences – limitations or restrictions See Zimbardo et al/ (1981) 21 22 Which is the independent variable? Which is the dependent variable? The purpose of this article is... ? Independent (predictor) variables Girls are less aggressive than are boys Math achievement is higher in boys than in girls ? Low income parents are more authoritarian than middle income parents ? Children in day care are less securely attached than home-reared children ? Couples who cohabit before marriage are more likely to get divorced ? – possible cause ? ? Dependent (outcome) variables – effect The independent (predictor) variable causes (??) the change in the dependent (outcome) variable 23 24 4 Two ends of the continuum A continuum of research methods ? Description ? Description Validation Validation – Experiments – Validation of explanation – Case studies – Exploration for explanation – Typically close to natural situation – Verbal description – Often a laboratory situation – Measurement and statistics – Structured, carefully planned data collection – Unstructured, spontaneous data collection – Rarely any treatment – Treatment administered to participants 25 26 Survey research: the swing point Description Survey Validation Case studies Surveys Experiments Exploration Either; both Validation Natural Rarely lab Special emphases of methods Description Explanation Laboratory Verbal Either; both Statistics Unstructured collection Planned or unstructured Planned & structured Qualitative XXXXX XXXXX Survey XXXXX X Single Subject XXX X Longitudinal XXX Validation X X Experimental X XXXXX Model building X XXXXX 27 28 Starting social science research: Ideas from Research is related to Theory Theory Research Literature ? Observations ? Serendipity ? Personal Experience ? ? Deductive research: ? Inductive research – start with the theory ? – Theory grows out of findings – “Grounded theory” 29 30 5 Theory Functions of a Theory Systematic explanation of observed facts and laws ? Relating to a particular aspect of life Describe ? Explain ? Predict ? Intervene If theory explains, it can predict and intervene ? ? 31 32 Characteristics of a theory Offers an explanation of some phenomenon Ties into the larger world ? Identifies critical and irrelevant factors (variables) ? Research Questions and Hypotheses ? 33 34 Values can be nominal, ordinal, interval, or ratio Basic vocabulary: Think of variables for each measurement level: ? Attribute or value ? – characteristic or quality that describes something – people or groups of people – e.g., male, female; mother, father ? Nominal values – Names, categories – No “more than, less than” meaning ? Ordinal values – More than, less than relationship – No equal distance between values Variable ? Interval ? Ratio ? Need more practice? See HD FS 503 Handbook – equal intervals between data points – logical grouping of attributes – MUST vary!! – True zero point 35 36 6 Research questions Find the variables and the attributes (values): ? ? ? ? ? The heart of description and explanation is the relationships between ? variables and ? attributes (values) Girls are less aggressive than are boys Math achievement is higher in boys than in girls Low income parents are more authoritarian than middle income parents Children in day care are less securely attached than home-reared children Couples who cohabit before marriage are more likely to get divorced 37 38 Hypotheses Research questions Specific expectations about empirical reality that are derived from theory ? Simplest form: ? Alternates to formal hypotheses – X causes?? (is associated with) Y ? More complex: – X causes U, which causes Z – Under condition W, X causes Y 39 40 Hypotheses dictate methods! Causal Modeling ? Methods do not dictate hypotheses! ? Tension exists between quantitative and qualitative researchers ? Guiding principle: ? Traditional model in social science research – searching for factors that explain the occcurrence of other factors ? Probabilistic – what is the probability that A causes B – Idiographic: » enumerates all causes – Nomothetic: » enumerates the most important causes: » probabilistic – Is it worth knowing? 41 42 7 Criteria for causality ? Necessary and sufficient causes Time order: ? Necessary: ? Sufficient: – dependent cannot occur without independent – cause must precede effect ? Susceptibility of influence – Dependent variable almost always occurs when the independent variable is present ? Goal in causal modeling: to identify necessary and sufficient causes 43 44 Design places boundaries around the research question Research Design ? ? Designation of important variables, Derived from hypothesis or research question ? ? Limits the population studied Statistical Controls – Gather data about relevant variables to control for alternate explanations ? Sample study 45 46 Unit of analysis ? ? ? Time frame Individuals? Families? Groups? ? ? – Therapy groups – Classrooms of children – Clients of a home visitor ? Cross-sectional Longitudinal – Cohort – Panel – Retrospective Critical issue BEFORE – selection of participants – entry of data 47 48 8 Before Tuesday Lab To study causes, you need comparisons! VARIABLES MUST VARY! ? At least 20% of participants must be in each attribute group ? If you’re studying the impact of therapy, you need ? Lab is in 108 Mackay ? Do NOT buy SPSS until you talk with me or Wick! ? ? – Grad pack only (~ $200), – Do NOT buy the student pack (no syntax mode) – Make sure your system has sufficient memory – one therapy and a control group – two therapies (i.e., demonstration and comparison) ? Consider the ethical issues here! 49 50 Review APA Style Next Monday Class ? Authors’ names: First name or first initial? ? First line: Indented or straight? ? Article: Capitalized? ? Journal name: Underlined or Italics? ? Volume number: issue number? ? Pages: pp.? ? ? ? January 28 (Enjoy MLK holiday!) E-mail me your learning team Check syllabus for assignments: – E-mail me your individual reviews of articles in Chapter 1 » Zimbardo, Andersen, & Kabut (1981) » Hoffman-Reim (1986) – 1.1 (use APA style for references) due Jan. 22 (hard copy) – 1.2 (use APA style for references) due Jan. 28 – Responses to questions in chapters 1, 2, 3, 4 ? 51 Bring/e-mail any questions about course organization, assessments, grading, etc. 52 January 28: 1.2 due Review assignment Check APA manual for references!! ? Self-assess using criteria sheet in Handbook ? Attach criteria sheet from Handbook ? Turn in one assignment for the two of you, after collaborating on the article ? Office Hours: Tuesday, 3 – 5 PM ? ? 53 9