O GoConqr está de visual novo! Conte para nós o que você achou.

Registre-se grátis

Flashcards

Methods in Management

Maxi Brausch Methods in Managemenet Alemão 104 flashcards

Flashcards do conjunto (104)

  1. 1Bootstrapping is useful when:- the number of oberservation is too Low
    - the asymtotic Distribution of an estimator is unkown
    - controlling for Data characteristics is required
  2. 2Characteristics of qualitative research:- Reflexivity
    - Context
    - Explication
    - Openness
    -Multiple perspections
    -Flexibility
    -Holistic account
  3. 3Qualitative thinking based on Mayring:- Description (1-3)
    - Interpretation (4-6)
    - Everyday Life & Subject (7-9)
    Generalization process (10-13)
  4. 4Phasen of discussion:1. Unfamiliarity
    2. Orientation
    3. Adaption
    4. Familiarity
    5. Conformity
  5. 5Types of Sampling:- Theoretical Sampling: based on concepts (grounded theory), qualitative
    - Convenience Sampling: random selection, not driven by theory, quantitative
    -Representative Sampling: e.g. group of 30, 15 m, 15f, hard to implement, quantitative
  6. 6Linear strategy (Quantitative research):Theory => Hypotheses => Operationalization => Sampling => Collection => Interpretation => Validation
  7. 7Circular strategy (Qualitative research):Preliminary Assumptions
    Selection of Procedure ->
    Sampling -> Collection-> Interpretation
    Theory Development
  8. 8Qualitative research cycle:Research Design
    Data Collection
    Sampling Strategies
    Types of Data
    From collection to analysis
    Data Analysis
  9. 9When to use qualitative research:- a complex, detailed Wunderstanding of the issue is needed
    - a desire to empower individual existis
    - a literarg, flexible style of Reporting is appropriate
    - a theory to Adress gaps in understandings is developed
    - a lack of fit between quantative measures and the Problem exists
  10. 10Variation of group-discussion:- homogeneous vs. non-homogeneous
    - artifically vs. natural
    - random vs. deliberate
    - structuredness vs. openess
    - Direktive vs. non-directive
  11. 11Evaluation Focus Groups:+ deep insight (thinking, feeling)
    + authenticity due to the language
    + variety of Topics / opinions
    +low economic effort
    + high interactivity
    - Prävention of individual
    - changes in opionions
    - silent participants
  12. 12Implementation of a Focus Group:1. Classification of ethical issues
    (Recording, using of Data, anonymization)
    2. Allocation of group members
    (Name tags + deskriptive Data, classifications)
    3. Assessment of Prior knowledge and understanding
    (Agreement on Termins, sharing of Experiments, Statement of interest)
  13. 13Steps in design and use of Focus Groups:1. Problem Definition / formulation of the research question
    2. Identifying the Participants
    3. Identification of Moderator and Design of the Interview Guide
    4. Recruitment of the Participants
    5. Conducting the group
    6. Analyzing and Interpretation of Data
    7. Writing the Report / Paper
    8. Decision making + Action
  14. 14Developing a research question:Research Problem Statement
    (Issue is need of investigation)
    Research Purpose Statement
    (Advance the Major Objective for beginning the study)
    Research Question
    (specifying the building Query for narrowing the study)
  15. 156 criteria to ensure qualitative research:1. Documentation of procedures
    2. Argumentative backing up of Interpretation
    3. Rule driveness
    4. Proximity to the subject under discussion
    5. Communicative validation
    6. Triangulation (Data, Methods, Theory, Investigator)
  16. 16Deductive category building:is based on already existing structuring of the object area in the literature
  17. 17Induktive theory building:is deriving the categories directly from the Data material
  18. 18Pro´s of Primary Data:+ exclusivity
    + control over Data collection
    + suitability
    + recency of Data
  19. 19Con´s of Primary Data:- economic disadvantages (time-consuming, costly)
    - requires Expertise in Data collection techniques
    - often Small sample sizes
    - biased Feedback
  20. 20Pro´s of Secondary Data:+ economic reason
    + Data Quality and legitimity
    + Larger sample size
    + complementing Primar Data
  21. 21Con´s of Secondary Data:- suitability
    - lack of control
    - lack of exclusivity
  22. 22Preprepatory Work:1. Choice & Evaluation of the research Problem
    (Pression of the Problem, feasability of empirical Examination, Scientific Scope)
    2. Theory Building
    3. Operationalization + Measurement
  23. 23Objectivity (Independence):Conduct
    Evaluation
    Interpretation
  24. 24Reliability (Consitence):Test-Retest
    Parallel-Test
    Inter-Rater
    Internal Consistence
  25. 25Reasons to conduct an additive Index:- unsystematisch Errors can be reduced
    - the mensuren theoretical concept concerns itself with several oberservable facts an circumstances, which is often the case in Business Administration Studies
  26. 26Cronbachs Alpha:- Exposés the Korrelation of all indicators among themselves (Intercorrelation)
    - can take values between 0 and 1, whereby the value of 0.7 is commonly considered as adequate
    - is rising when increasing the number of indicators
  27. 27Validity:Content (not measurable)
    Criterion (Prognostic or Conformity)
    Construct (Convergence or Discriminant)
  28. 28Convergence / Discriminant Validity:Convergence: Do two different constructs on measuring "Business Strategien" arrive the same conclusions?
    Discriminant: Does the measuring instrument measure different than others would do?
  29. 29Internal vs. External Validity:Internal: a dependency Problem (statistical Conclusion)
    External: a generalization Problem (Ecological Validation)
    Trade-Off: A tight controlled research setting provides high internal validity, a Realistic setting provides high external validity
  30. 30Confounding factors Internal Validity:History (Time between pre/posttest)
    Maturation (biological / psychological)
    Measurement Effects (changes in the dependente variable)
    Instrumentation (Adaption in Survey)
    Statistical regression (extreme values)
    Selection (Bias)
    Mortality (Dropouts)
  31. 31Additional confounding factors for Field Experiments:Resentful Demortalization (CG does not receiver a desirable Treatment)
    Compensatory Rivalry (CG tries to perform as EG eventough they have disadvantages)
    Compensatory Equalization (CG becomes a Kind of Treatment)
    Treatment diffusion( CG tries to anticpate their reactions)
  32. 32Confounding factors fo External Validity:Lack of instrumental validity
    Sample Error
    Experimental Reactivity
    Pretest Effects (Single Out / Interaction)
    Hawthorne Effect
  33. 33What to do regarding confounding factors of external validity:Elimination, Keep Constant, Register
    Personal confounding: Randomization, Keep constant, Parallelization, Matching
  34. 34Experiments- are tested Arrangements to verify hypothesis on casual relations
    - specific Variation of independent variables
    - Observation of dopendent variables
    - Elimination of confounding factors by controlling the Experiment setting
  35. 35Advantages over non-experiemnts+ potential to influence factors
    + Sole Method to prove casual relations
  36. 36Pro´s of non-experimental Data (e.g Surveys)+ economic reasons
    + large Data Sets
    + ease of generalization through choice simplicity of representative samples
  37. 37Data Collection Procedure- Cross section (same research obejcts, 1 Point of time)
    - Pooled Cross section (different r.o, different points of time)
    - Time series (same r.o, in a chronological Order)
    - Panel (multiple sources, large period)
  38. 38Pro´s & Con´s of Surveys:+ cheap, huge number of participants, less biased, often high anonymity
    - Low represtativeness, difficult to Formulare questions, technical implementation, anonymity: fakes answers
  39. 39Pro´s & Con´s of Quantitative Data (closed-eded questions)+ good collectability, comparability, evaluability
    - loss of info, cant react to Problems -> time consuming, closed end: social desirability
  40. 40Pro´s & Con´s of Qualitative Data (open-endet questions)+ richer Data, react to problems, unbiased
    - difficult to Measure, compare and evaluate, smaller spample sizes, long post-treatment, participants might have Problems to artivuating themselves
  41. 41Planning an oral Survey:- Setting
    - Micro planning (determinier subject areas)
    - Marco planning (formulating the individual questions)
    - Survey technique (sequencing of questions)
    - Interviewer training
  42. 42Setting of an oral Survey:- Authority requirement of the Interviewer (ephatic to aggressive)
    - Type of conduct (direct or indirekt)
    - Number of Interviewees (1:1, group)
    - Nummer of Interviewers
    - Function (Investigating or conveying)
  43. 43Degree of standardization:Structured (wording & ordering is fixed)
    Semi-Structured (Framework, a bit wording/ordering)
    Unstructured (Only Framework)
  44. 44Pro´s & Con´s of semi-structured:+ social desirability effect mostly eliminated, getting more detailed & profound insights, softer construct, rich data still quantifiable
    - Responses might be biased, no: double scoring, double Interviews
  45. 45Organizational Design Elements:Task Allocation
    Planning and controlling systems
    Incentive System and compensation
    Information System and Communication
    Talent Management
  46. 46Pro´s & Con´s of pre-existing Databases:+ less time-consuming, Data breadth, Data Quality
    - lack of exclusivity, expensive license, lack of Data death, lack of Data knowledge
  47. 47Pro´s & Con´s of Datasets:+ large sample size, suitability, exclusivity, maybe Low costs
    - high costs if you have to Buy it, dependency on political issues, lack of Data knowledge, Accessibility
  48. 48Manual Data Collection:1. Conceptual Work: consulting initial Desk research, literature review, operationalize construct of Interesse, defining the sample
    2. Data Collection: actual Data Katherine
    3. Coding and Measurement: developing a systematic Coding scheme, acutal Coding, further robustness checks
  49. 49Pro´s & Con´s of Manuel Data Collection:+ exclusivity, suitability, in-depth knowledge
    - high aquistions costs, smaller sample size, prone to Errors, validity issues
  50. 50Data Crawling:= large scale extraction of Information from the internet
    useful when: too many observations for Manuel Data collection are, Data allows craling, no login is required
  51. 51Pro´s & Con´s of Data Crawling:+ exclusivity, fast collection Methoden, suitability, Independence of third Parties, large sample sizes
    - learning costs, Prone to Errors, lack of Data knowledge
  52. 525 Steps of formulating definitions & Coding rules:1. Determination of the units of analysis
    2. Establishing assesment dimensions
    3. Determinist the values
    4. Definitions, anchor samples + encoding rules
    5. Markung of points of discovery
    6. Assignment of categories
  53. 53Requirements of category Schemas:1. content-related
    2. Selectivity
    3. Completeness
    4. Unambiguity
  54. 54Experimental Designs:- True (projective) Experiment: Active Variation of the independent variable by the experimenter
    - Ex-post-facto-Experiment: Comparsion of 2 natural groups which Differ in regards to the independent variable
    - Solomon-4-group
    - Reverse-treatment
  55. 55Pro´s & Con´s of a true experimental design:+ Direct comparasion of EG and CG possible
    - bias, interpendencies
  56. 56Randomizationnot to harm i.v., Stops outside influcence on EG/CG, equalizing confounding factors
    Limits: precondition, Problems with ex-post-facto-exp., comparability at Start might dimish over the course of the test
  57. 57Pro´s & Con´s of Solomon-4-Group:+ all confounders can be controlled (i.v), e.v is determinable, control of pretest effects, estimation of the unbiased Treatment effect is possible
    - huge effort (time, money)
  58. 58Pro´s & Con´s of Reversed treament:+ higher contruct validity than without CG, Hawthorne effect gets migrated
    - no conclusions, often a neg. Treatment isnt applicable (e.g smokers)
  59. 59Field vs. Labatory ExperimentNatural setting vs. artifical setting
    disclosure vs. Nondisclosure
    external vs. internal validity
  60. 60Empirical Research is influenced by:- Consumer insights
    - Competitive Advantage
    - Communications Management
  61. 61Reputation is influenced by:Likeability (Attractiveness, Corporate Social Responsibility)
    Competence (Quality, Performance)
    Corporate Communications
    Costoms, Employees, Investors etc
  62. 62Corporate Communications / Quality:Products / Services
    Strategy / Innovation
    Manager / Employees
    Performance / Market Position
    CSR / Fairness
  63. 63Brand Equity is determined by:- Awareness (Uniquness, Degree of popularity)
    - Quality (Functionality)
    - Loyalty (First choice, preference)
    - Image (Likeability, Attractiveness, overall attidude)
  64. 64Pro´s & Con´s of Hazards:+ controlled Survey environment due to labatory exp., controlling for confounding factors, CG in Order to observe measurement bias
    - experimental reactivity, student sample
  65. 65How to increase external validity:- Additional Field Experiments
    - Additional product categories
    - larger sample size
  66. 66Pro´s & Con´s of Student Samples:+ easy recruitment, Low costs, no monetary reward necessary, homogenous test sample (high i.v.)
    - not representative for the whole Population, Low e.v.
  67. 67Problems with traditional literature reviews:- defined to broadly, fuzzily
    - research State May no completely be retrieved
    - clearly defined criteria isnt guaranteed
  68. 68Systematic reviews:- criteria for study Retrieval and selection of Studies, and for Data Extraktion are clearly stated
    - results are dran from Studies using a traceable analyitcal process
    - the State of research is completely covered
  69. 69Aims of systematic reviews:- comprehensive, complete analysis of the existing State research on a well defined problem
    - creation of an evidence Basis the Quality of which is made transparent & used for decision making
  70. 70Workplan for a Systematic review:1. research question, review aim
    2. Studies to be included, criteria to be assesed
    3. literature review
    4. study selection
    5. study assessment
    6. Synthesis of the result
    7. Report, dissemination
  71. 71Limiations of a systematic review:- Search Strategies (not comprehensive)
    - Criteria to assess validity (inadequately reported)
    - Methods of synthesizing results (not reported adequately)
    Conclusions ( not based on review Findlings)
  72. 728 questions to check a systematic review:+ PICOS (Population, Intervention, comparators, outcomes, Study designs)
    + search strategy
    + how to minimize Errors/bias
    + appropriate criteria?
    + adequate Details?
    + appropriate Methods?
    + do conclusions reflect the evidence?
  73. 73Quality of Meta-Analysis:- the choice of the study endpoints
    - control of measurement Errors/outliers
    - analysis of homogenity and eventual Publikation bias
    - Limitation gettings better with more recent Studies
  74. 74Meta-Analysis are strong when:- similiar Quant. research existits
    - the Integrale of empirical Data & results
    - go beyond conclusions of Single Author teams
    - evidence-based generalizations across studies
    - reject conclusions from Studies that havent been retrieved systematically
  75. 75Narrative Reviews are useful when:- to Elaborate study nuances
    - to Report path-breaking research
    - to synthesize losely connected research topics
  76. 76PRISMA= Preferred Reporting Items for Systematic Reviews and Meta-Analysis
  77. 77Flow Chart of Prisma:1. Identification
    2. Screening
    3. Inclusive and exclusion criteria
    4. Studies included
  78. 78Scoping reviews:- exploring contributions in eventually emerging, broadly defined research Fields with different study design/Quality
    - defining research Fields with erhard to types of Topics covered
    - preliminary work and feasibility test for a systematic review
    - analytical summary of the State of research for desicion makers
    - identification of research gaps
  79. 79Using Models if:- Experiments are not feasible
    - Problem is too complex for primäry study
    - use of existing knowledge
    - Exploration as the aim
  80. 80Models should be:- Simple (easy to Understand)
    - Robust (not rendering absurd results)
    - Easy to use
    - Flexible (can be updated easily)
    - Complete (reports important facts)
    - Easy to communicate
  81. 81Pro´s & Con´s of Models:+ long-term analysis, cheap, strategy comparision simple, best available evidence can be used, real word modelling possible
    - arbitrainess, einproper Extrapolation, Quality difficult to assess, transparent May be Limited
  82. 82Pro´s & Con´s of Experiments:+ Quality, no bias, long Standing, substianial empirical answer to the research question
    - costly, answer only one question, time Horizont restriced, cause-effect not real-world
  83. 83Use in health economic Evaluation:- assessment of cost-effectiveness
    - development & use of a Prävention model
    - probalistic sensitive analysis
    - use of modelling Standards
    - experience from modelling
  84. 84Pro´s & Con´s of Specific Software+ many Model Variante, graphical representation, long-year Testing, learning support
    - blackbox, time lag for innovative modelling Methode
  85. 85Pro´s & Con´s of Programming Software+ outside pre-programmed forms, full access to Model Programm, Data -> to new needs, flexible programming
    - more risky for bugs, graphical representation difficult
  86. 86Dokumentation of Models:Type
    Process
    Testing (Validity)
    Technique
  87. 87Markov Models:finite number of disjunctive States which are passed through in "n"-cycles, Transition probabilities between States are Constant troughout all n cycles
    => no Interaktion of individuals, cohorts, sensitivity analyses
  88. 88Decision Tree Models:decision Nodes Reflex strategy choices, finite number of disjunctive pathways leading to an end node, one cycle one, probabilites at Chance Nodes are fixed
    => no Interaktion of individuals, cohorts, sensitivity analyses
  89. 89Pro´s & Con´s of Markov Models:+ perform well in long time Horizonts ( chronical diseases, screenings programms, complex problems)
    - comparatively complex
  90. 90Pro´s & Con´s of Decision Tree Models:+ simple (short term Horizonts, clinical Studies, disease preventation, calculate long-termin costs on the Basis of empiric short-vermin costs)
    - very complicated in long-time Horizonts
  91. 91Models are good if:- it enables the Synthesis of Different Data and results
    - perkins quantification of uncertainty
  92. 92Problems of Models:- inadmissable Extrapolation
    - Manipulation of data
    - assumptions bias Model results
    - ensuring Quality is difficult
    - transparency and comprehensibility are hard to establish
  93. 93Prevention of problems:+ extensive, precise and transparent description and all paramenters
    - e.v. & i.v.
    - extensive sensitivity analsis
  94. 94Sensitivity Analysis:Basic assumptions lead to baseline result, sensitivity analysis change these assumptsion and May change the results
    => motivalte Parameter Rangers investigativ and types of analyses used
  95. 95Types of Sensitivity Analysis:Deterministisch (One-Way-A.) - Best Case / Worst Case
    Probalistic (One-Way-A.) - all random variables
  96. 96Phases in research Processes (CIT´s)1: Problem Definition
    2: Study Design
    3: Data Collection
    4: Data Analysis / Interpretation
    5: Report Results
  97. 97Pro´s & Con´s of CIT+ pure consumer Data, provides concrete Information, allows to explore a specific Phänomen, well Suite for use in assessing perceptions of custom from different cultures
    - issues related to reliability/validity, bias, negative incidents, economics reasons
  98. 988 Basic principles of qualitative content analyses:1. Embedding of material in Communication context
    2. systematic rule-based approach
    3. categories are Central to the analysis
    4. inclusive of object rather Focus on technique
    5. review of specific Instruments
    6. analysis is theory-driven
    7. involvement of quantitative steps in the analysis
    8. Quality criteria
  99. 99Basic forms of Interpretation:- Summary (Inductive)
    - Explication
    - Structuring (Deductive)
  100. 100Monte-Carlo-Simulation:= a procedure to numerically solve problems by repeating random Sampling many times, highly useful if the Problem cant be solved analytically
  101. 101Applications of M-C-S:- analyze the Properties of statistical estimators
    - Evaluation of Financial assets
    - Scenario analyses (if high stocastic Input and complex Dependence structures)
  102. 102Steps of M-C-S:1: Model (true)
    2: Simulation (Sampling)
    3: Estimation (Use Data of 2, to calculate the variable of interest)
    4: Repetion (Repeat 2+3)
    5: Summary
  103. 103Bootstrapping is useful if:- the number of observations is too low
    - the asymptotic Distribution of an estimator is unknown
    - controlling for Data characteristics is required
  104. 104OLS assumptions for the linear Model:1. Population orthogonality assumption
    2. No Mulitcollinearity
    3. Homoscedasticity

Informações

Descrição
Methods in Managemenet FlashCards sobre Methods in Management , criado por Maxi Brausch em 19-02-2018.
Criado por
Maxi Brausch · fev 2018
Atualizado
19 fev 2018
Visualizações
2
Avaliação

Fale Conosco

Sobre o que é sua mensagem?

O que você acha do novo visual?

Você gostou?