Modeling nature Task 1 Theory construction and Model-building skills (chapter 1-3) Scientific research is a process that is designed to extend our understandings and to determine if they are correct or useful. Science strives to operate according to a more rigorous set of rules. Realism: reality exists independent of any human presence, there is an external world comprised of objects that follow a myriad of natural facts and laws. –> mirror world. It is possible to make claims regarding ultimate truths.

Social constructionist: reality is a construction of the human mind, which is tied to a particular time and social context, reality changes as the social context changes. There is no reality and there are no facts until these are conceptualized and shared by some number of people Critical realism/ Blumer: reality is seen through human conceptions of it, but the empirical world also talks back to our conceptions in the sense of challenging, resisting, and failing to bend to them. This inflexible character of the world justifies empirical science.

Hypothetical realism: even if it cannot be proven that reality exists, it is useful to assume that is does. It is a heuristic device: helps to organize our thoughts and accomplish goals and objectives. The way in which reality is interpreted can vary within the individual over time, across individuals, and can be heavily influenced by context. Reality appears: –> refers to the external and internal environment complex dynamic; things never stay the same unique; because of the dynamic nature the universe is never the same obscure; the vast majority of nature remains hidden from direct view

Concepts are the fundamental building blocks of everyday life as well as from scientific thinking. Concept refers to something that is conceived of in the mind, a generic idea or thought, usually developed from experiencing one or more particular instances. basic level of understanding is identification generalized abstractions, general idea that can be applied across a number of specific instances encompass universes of possibilities, each concept consists of a universe of content hypothetical, not reality just ideas regarding reality , but the things they refer to are observable entities as well as non-tangible phenomena. re mostly learned, but some ay be hardwired (genetically determined) socially shared, in order to communicate reality orientated (or functional), used as a guide for interpreting and reacting to the world selective constructions, concepts are applied to describe reality depending upon the needs and objectives of the individual conceptualizing Variables make up concepts Constructs: higher order of concepts, referring to instances that are constructed form concepts at lower levels of abstraction. Powerful means by which we are able to handle greater portions of reality. Variable: composed of different levels or values, focus of social science.

Important because people and social entities are thought to differ depending on the variable category. Process-oriented theories: rely more on process- oriented characterizations of phenomena and/or on narratives. It is only when concepts are placed into relationship with each other that they move us toward achieving a deeper understanding of our reality. Different forms of relationships are: spatial, temporal, deterministic, legal. –> two concepts are linked we have conceptual system, which enables us to arrive at deeper levels of understanding. Selection mechanisms: the nature of the conceptual system that is invoked epends upon the needs of the individual at that moment. Scientific theories are essentially conceptual systems designed to be useful in identifying, organizing, and explaining or predicting some delimited portion of the experienced world. Features of understanding: Explanation: core facet of a conceptual system derived to provide insights into a phenomenon. predicting differentiate between concepts and events identifying describing organizing Understanding gives the ability to control events or relationships involving two components understanding the relevant features of the environment aving the ability to manipulate those features Communication: a process whereby a source transmit a message over a medium to one or more receivers. Meaning structure: concepts of thoughts that exist in the mind of individuals Surface structure: symbols that are the externally visible expression of these thoughts Meanings can be expressed via a universe of possible symbols, since they can comprehend different meanings communication is enhanced by careful attention to the selection and use of symbols. Communication requires that the receiver has a concept comparable to the one in mind of the source.

The existence of shared symbols enables us to communicate, revise thinking, clarify logic, tap into the accumulated wisdom of the past. Shared meaning systems. Three fundamental characteristics of shared approaches Each consists of concepts and relationships among these concepts shared systems tend to be more elaborate, more abstract, stabler over time, and more explicit than individual conceptual systems limited in how much of the world they address generally serve prescriptive (guidance for how to approach or respond) and evaluative functions (permits labeling something)

Theories need to be useful, provide us with a useful way of describing or coping with the world about us. Consensual validation: the worth of a particular conceptualization is gauged by the degree of acceptance it is granted by others. Expert validation: the decision whether a particular conceptualization merits acceptance is determined by selected others who presumably have the knowledge and wisdom to discern what is and is not correct. Internal validation: involves the application of formal rules of logic to examine the concepts within a particular conceptual system.

After intensive logical assessment conceptualization is said to be confirmed. Systematic empirical validation: scientific conceptualizations tend to be accepted only to the extent that they have been subjected to rigorous and systematic empirical testing and shown to be useful. Science conceptual realm: entails the development of a conceptual system that can be communicated unambiguously to others empirical realm: the process whereby the worth of the conceptualization is assessed through the conduct of scientific studies Conceptual systems are prescientific, to be scientific they need to be subjected to empirical testing.

Metaphysical explanations: conceptualizations that cannot be publicly observed and tested. Theoretical propositions must be addressed in the empirical realm for science to progress. The emphasis of empirical (dis)conformation and the process by which this is accomplished are the sine qua non of science and distinguishes it from all the other approaches generating understanding. Other validations used but not seen as scientific. Theorizing/ theory construction: the process of formulating conceptual systems and converting them into symbolic expressions Theory: set of statements about the relationship between two or more concepts or constructs.

Models: involve concepts and relationships between concepts, mostly interchangeable with theory Hypothesis: empirically testable statements that are derived from theories and that form a basis for rejecting or not rejecting those theories, depending on the results of empirical testing. Involve concepts and relationships between them Theoretical expression/ conceptual systems: theories, hypothesis, models Applied research: Research that focus on an immediate problem, relies on concepts that are relatively narrow in scope, and produces results that are not intended to extend a general body of knowledge.

Basic research: conducted for the purpose of extending the boundaries of our collective body of understanding, not for the basis of solving a problem. Occupied with broader, less concrete concepts. Characteristics of a good theory utility: useful guides to the world we experience Shaw and Costanzo accepted in the scientific community logically consistent in agreement with data and facts testable desirable but not essential to acceptance clear and understandable terms strive to be parsimonious; adequate explanation, simple onsistent with other accepted theories that have achieved consensus scope: the greater the range the better (controversial) creative, novel generate research activity A scientist brings to any setting a prior schema that is used to filter, interpret, and analyze the world about him or her, which influence the research. intersubjectivity: the fact that other scientist agree on the empirical existence and could reproduce them Science attempts to yield perspectives on the obdurate character of our social and physical environment, in an unbiased and unprejudiced manner. Task 2

Theory construction and Model-building skills (chapter 7) Causal modeling: people are conceptualized as varying on some construct, theorist are interested in understanding what causes this variation. Focus on cause-effect relationships Causal thinking: tries to explain variability by identifying its causes. When something causes variability then it that something also varies. Causal analyses: involves identifying relationships between variables, with the idea that variation is one variable produces or causes variation in the other variable. Involves specifying effects of variables.

Predictive relationship: focus on the question is variability in A related to variability in B? If so we can predict B if we have knowledge of A. Focus on prediction rather than causation. Causal relationship: relationships invoke the notion of causality, with the idea that one of the variables in the relationship, X, influences the other variable in the relationship, Y. Causality: an elusive concept that is fraught with ambiguities. Can rarely be demonstrated unequivocally. Concept is a heuristic. By thinking in causal terms we are able to identify systematic changes in phenomena that are scientifically or socially desirable to change.

Scientific research is conducted to establish strong, moderate, or weak levels of confidence in theoretical statements that propose causal relationships. Russell: causality can be established unambiguously only in a completely isolated system. (How can we now it is a completely isolated system? ) When contaminating variables are present it is possible for a true causal relationship to exist even though observations show that X and Y are completely unrelated to each other. As the other way around. Common features If X causes Y, than changes in X are thought to produce changes in Y a cause must always precede an effect in time he time that it takes for a change in X to produce an change in Y can vary , ranging from virtually instantaneous change to year, decades.. cause and effect must be in some form of spatial contact or connected by a chain of intermediate events nature and strength of the effect can vary Causal realism: phenomena within the objective world are so intertwined and so dependent on one another in such complex ways that simple variable-centered notions of causal regularities are inadequate. Independent variable/ determinant: presumed cause

Dependent variable/ outcome variable: presumed effect 6 types of causal relationships: (Figure 7. 1) direct: in which a given cause is assumed to have a direct causal impact on some outcome variable. indirect/ mediated relationships: a variable influences another variable indirectly through its impact on an intermediary variable spurious relationship: in which two variables are related because they share a common cause, but not because either causes the other moderated: the causal relationship between variables, X and Y, differs depending on the value of a third variable, Z. idirectional: when two variables are conceptualized as influencing each other unanalyzed: two variables are connected, but the theorist is not going to specify why they are correlated. Path diagram: a figure where variables are indicated by a box and causal influence is represented by a straight arrow emanating from the cause and pointing to the effect. Multivariate causal model: model with more than one of the six types in them Endogenous variable: has at least one causal arrow pointing to it Exogenous variable: does not have a causal arrow pointing to it

Construction a causal theory: Identifying outcome variables: Choose a outcome variable you want to explain specify the variable and ask what are the consequences or effects of it build a theory on the effects of an intervention on an outcome Identifying direct causes: specify two or three variables that are direct causes of the outcome variable, with the goal to explain why there is variation in the outcome variable by choosing an initial variable but treated as a cause rather than an effect, identify two or three variables that the variable is thought to impact.

Turning direct causal relationships into indirect relationships through the specification of mediators: identify variables that mediate the direct relationship (why heuristics) Complete mediator: Z completely mediates any impact X has on Y Partial mediator: Z only partially mediates the effects of X on Y (Is there some other mechanism that Z by which X influences Y? ) Turning direct causal relationships into indirect relationships by treating direct causes as outcome variables = cause of a cause heuristics Moderate causal relationships: involves 3 variables, a cause (X), an effect (Y), and a moderator variable (Z).

Essence of a moderate relationship is that the strength or nature of the effect of X on Y varies as a function on Z. Carefully examine the circumstances you generate and try to abstract a variable that captures or represents them. mediated moderation partial mediated moderation moderated mediation moderated moderation Consider adding a mediated moderator relationship. Moderate mediation occurs when the strength of a path signifying mediation varies as a function of some variable, Q. Q also might moderate the moderated relationship (second- order moderator variable Consider introducing reciprocal causality

There is no such thing as simultaneous reciprocal causality, because there always must be a time interval, between cause and effect that follows from that cause Feedback loops: X influences Y which influences Z, which in turn feed backs to influence X. –> reciprocal causal relationship with a mediator variable inserted into to causal chain (use why? heuristics) Moderator variables can be added to one or both of the reciprocal causal paths (stronger than? heuristics) (cause heuristic, what causes it? )

Additional steps to be considered (may create spurious relationships) Adding additional outcomes: consider variables that are conceptually related to your initial outcome variable. If so, you must specify how all of the variables in your theory are related to them by adding appropriate causal paths Adding effects of effects: make your effect a cause –> turns initial effect into a mediator, consider mediate moderator, moderate causal relationship Specifying causal relationships between existing variables: map out causal pathways, by creating new (in) direct effects consider the tools.

Recognize that the exogenous variables are correlated. The relationship is only mentioned when there is a strong theoretical reason for saying there is zero correlation between them. Draw curve two- headed arrows among al exogenous variables or note at the bottom that al exogenous variables are assumed to be correlated. Details you should consider Temporal dynamics Three types:

Contemporaneous effects: the effect of X on Y within a given time period Autoregressive effects: where a variable at one point in time is assumed to influence a person’s standing on that same variable at a later point in time Lagged effects: effects of a variable at time 1 on the other variable at time 2, independent of the contemporaneous and autoregressive effects that are operating. Choice of time intervals: should carefully think it through and have a rational for the intervals upon which you ultimately settle. Disturbance terms: presence explicitly recognizes that not all causal influences on a variable have been specified.

Only endogenous variables have this. Provide explicit statements about which disturbance terms in the framework are correlated and which disturbance terms are not. Longitudinal component: two point in time must be correlated. Measurement theory: makes a distinction between: (relation usually linear, but could also be non linear) latent variable: true construct that you are interested in making statements about Structural model: diagram focused on the causal relations among the latent variables observed measure of that variable (e= measurement error)

Measurement model: diagram with arrows from the latent constructs to the observed measures Revisiting your literature review Final steps: take every direct relationship you have and try reversing its sign show your theory to family and friends Path diagram: summarizes many theoretical propositions efficiently, that if expressed verbally, would constitute a long list. Every causal path represents a theoretical proposition, and the absence of casual paths also reflect theoretical propositions. = a positive relation = inverse relation

Nonlinear relationships can be described in the text, either verbally or mathematically Structural equation modeling (SEM): tends to be used with research design that are correlational or observational in nature rather than experimental, though SEM is also readily applied to experimental data. Each path in a diagram is a thesis or potential conclusion and the theorist needs to build as strong as case as possible for the underlying a priori logic. Grounded/ emergent theorist can use path diagrams as well. Theory construction and Model-building skills (chapter 8)

Mathematical modeling: describing relationships between variables, using mathematical concepts, based on functions. Category variable: has different levels, values, or categories and there is no special ordering to the categories along an underlying dimension. The categories are merely labels that differentiate one group from another. Quantitative variable: individuals are assigned numerical values to place them into different categories, and the numerical values have meaning in that they imply an underlying dimension that is of theoretical interest iscrete variable: there are finite number of values between two values continuous variable: there is an infinite number of values between any tow values Social scientist must rely on discrete measures of continuous constructs, they build models with them as if they were continuous Axiom: a mathematical statement that serves as a starting point form which other mathematical statements are logically derived. Axioms are given. Theorem: a statement that can be logically derived from, or is proven by, one or more axioms or previous statements.

Functions: involve numbers as inputs and outputs Domain: set of possible input values, any number that produces a meaningful output Range: set of possible output values Slope: indicates the number of units that variable Y changes when variable X increases by 1 unit. Describes the linear relationship. b= (Y2-Y1)/(X2-X1) positive slope: indicates a positive or direct linear relationship, as scores on X increase, scores on Y also increase negative slope: indicates a negative or inverse linear relationship, as scores on X increase, scores on Y decrease

Intercept: the value of Y when X is zero. Linear function: f(X)=a+bX Disturbance/ error term: accommodate random disparities Y=a+bX+e –> where e is the difference between the observed Y score and the predicted score based on linear function. Errors are random. E is an unmeasured variable that reflects the disparity between scores predicted by the model and observed scores. Deterministic model: there is no random error operating Probabilistic model/ stochastic model: in which some degree of randomness is present

Adjustable parameters/ adjustable constants: constants such as the intercept and the slope, their values can be set by the theorist to different values so as to affect the output of the function. Can be estimated empirically based on data. Fixed: when the value of an adjustable parameter is specified a priori by the theorist and not estimated. Estimated: when the value of the adjustable parameter is estimated form data. Derivatives: refer to the concept of instantaneous change

Differentiation: refers to algebraic methods for calculating the amount of instantaneous change that occurs. Rate of change/ derivative: (Y2-Y1)/(X2-X1) or dY/dX –> the change in Y relative to the change in X as the change in X approaches its lower limit of zero. Or instantaneous slope. Integral: reflects the amount of something. Area under the curve (normal distribution) between two points. Calculating it is called integration. Used to characterize accumulations: how much of something is accumulated.

Model identification: cases where the values of model parameters must be estimated from data. Just- identified model: one for which there is a unique solution for the value of each estimated parameter in the model Under-identified model: one for which there is an infinite number of solutions for one or more of the model parameters –> unsatisfactory Over- identified model: for which there is a unique solution for the model parameters, and there is more than one feature of the model that can be used to independently estimate the parameter values

Theorist give careful consideration to the metric they use to measure the variables, because the accuracy of a mathematical model and the inferences one makes can be influenced by the metric of the variables. Concavity: whether the rate change on a curve is increasing or decreasing concave upward: curve has an increasing first derivative concave downward: curve has a decreasing first derivative. Proportionality/ constant: two variables are proportional to one another when one variable is a multiple of the other. Y=cX –> c = constant inversely roportional: if two variables have some constant c for which Y=c/X Scaling constants: refer to adjustable parameters in a model that have no substantive meaning but are included to shift a variable from one metric to another. Six classes of models: Linear model: two adjustable parameters, a slope, an intercept, that typically are estimated rather than fixed by the theorist. Logarithmic functions: f(X) = loga(X), where a is a constant indicating the base of the logarithm. Used to model growth or change when the change is rapid at first and then slows down to a gradual and eventually almost nonexistent pace. ncrease by constant intervals over input multiples Log base 10 of 100 is log10(100), Natural log (Ln): uses a constant called e as its base, e’s value is fixed. Undefined for negative values value of the log can be positive or negative (a between 1 and 0) as the value of X approaches zero the value of the log approaches negative infinity when x=1, the value of the log is 0 as X approaches infinity the log of X also approaches infinity Exponential functions: f(X) = a^X, yields an output that increases in value with increasing x if a > 1 and decreases in value with increasing X if a is between 0 and 1.

Often used to refer growth or change with certain properties. Inverse of log functions, two functions mirror image each other’s properties. Increase by multiples over constant input intervals a between -1 and 0 output value is damped oscillation as X increases a is 1, when a > 1 the curve will be concave upward positive X > 1, when a between 0 and 1, curve will be concave downward Polynomial functions: f(X) = a + bX1+ cX2+ dX3+ . . . where X continues to be raised to the next highest integer value, and each term has a potentially unique adjustable constant.

Can model data with many wiggles and turns, the more wiggles the more terms that are required to model it. Adjusting one term to a linear model allows the model to accommodate a curve with one bend. Add up another and a second bend will occur. Quadratic (3 terms, U- shape) and cubic (four terms, S- shape) functions most popular. Trigonomic function: used to model cyclical phenomena. Most used:f(X) = sin(aX) and f(X) = cos(aX), where a is a constant, sin is the sine, and cos is the cosine.

Transformations of the sine and cosine functions can reproduce many forms of periodic behavior. Mathematical modelers select functions: a priori, based on logic after collecting data and scrutinizing scatterplots Mathematical modelers can create quantitative representations of categorical variables and then analyze the quantitative translation using the quantitative functions. Sometimes represented in a table or graph rather than a function. Rational functions: divides one polynomial function by a second polynomial function rather than summing polynomials.

Function transformation add/ subtract adjustable parameters to them Vertical shifts: shifting the output value upward or downward Vertical stretches (a > 1)/ crunches (a < 1): through multiplying the parameter a, before the rule described is applied Horizontal shifts: shifting the graph left (adding positive value) or right (subtracting a positive value) Horizontal stretches (a > 1)/ crunches (a > 1): multiply X before the rule described is applied Fine tune the form of the curve to your problem. Combining functions ogistic function f(X) = c/(1 + ae–bX) combines exponential and bounded exponential growth functions, resulting in a S- shaped curve special case of sigmoid function, which generates curves having an S- shape. World population growth can be measured. Break the overall relationship into smaller series of smaller component segments, specify a functions to reflect each segment, and then assemble the components functions into a larger whole. Multiple variable functions: functions with more than one input Building a mathematical model:

Identify the variables that will be included in the model and identify the metrics on which the variables are measured. Think carefully about the variables, the metrics, and the relationships between the variables and pose a few candidate functions that might capture the underlying dynamics. Implications, predictions. A working function is settled upon, typically a function that includes several adjustable constants. Collect empirical data, estimate values of the adjustable constants from data, examine the degree of fit between the output values and the function and the values observed in the real world.

If function not sufficient another must be tried. revise the model, apply model to new set of data to determine thow well the revised model performs. Chaos theory: refers to disarray, but what appears chaotic sometimes has a systematic base, or systematic function generating it. Requires precise measuring and can be influenced easily. t + 1 refers to the time period following t. difference equation: in case where a variable at time t is a function of variable at time t – 1. first- order difference equation: if the variable at time t is a function of the immediately preceding point in time second- order difference equation: t- 2 emporal chaos discrete time models: focus on discrete time intervals continuous time models: use time continuously spatial chaos: properties of space and distance are used Catastrophe theory: uses seven fundamental mathematical equations to describe discontinuous behavior. relates outcome variables to control variables (explanatory variables) relationship expressed mathematically using nonlinear, dynamic systems that rely on different forms of polynomial functions Catastrophic event: an event where a large and rapid change in a ystem output occurs even though the system inputs are smooth and continuous. Bilinear: assumes that the relationship between the outcome and the predictor is always linear no matter what the value is of the other predictor Theory construction and Model-building skills (chapter 4-6) Theory construction: involves specifying relationships between concepts in ways that create new insights into the phenomena we are interested in understanding. By invoking concepts and precesses that we think influence it or are the basis for it. enerating ideas about new explanatory constructs and relationships, underlying mechanisms and trying to explain the merits of these ideas more carful analytic scrutiny choose key constructs and relationships to focus upon, refine them Creativity: the ability to produce work that is both novel and appropriate 50s/ 60s: focus on lives and minds: Barron: creative person has typical characteristics Creative contributions are in the interaction of three systems: innovating person substantive domain in which the person works the field of the gatekeeper and practitioners who solicit, discourage, respond to, judge, and reward contributions.

Different disciplines describe creativity differently Creative ideas: provide novel perspectives on phenomena in ways that provide insights not previously recognized. crowd pleasers leading the crowd by accepting their presuppositions and analyzing next steps in thinking and by reaching those next steps before others do crowd defying ideas: eschew the presuppositions on which a body of knowledge is based Creative process Amabile: three core facets; high motivation domain- relevant knowledge and abilities to address the task creativity relevant skills

Sternberg et al. : six resources: knowledge, styles of thinking, personality, motivation, environment, intellectual abilities; (see problems in new ways and escape conventional thinking, see which ideas are important, persuade ideas) Think both globally and locally. –> be knowledgable about a field but don’t let it channel your thinking too much Simonton: creative ideas often happen randomly –> but for making an impact one need exceptional logic and intellect to recognize the underlying connections Management: seven stages for groups and individuals rientation: become familiar with the problem preparation analysis: consider as many perspectives as possible ideation incubation synthesis evaluation Consider communication strategies as well as idea- generation strategies. Start with the proper mindset, make the decision to be creative. Choosing what to theorize about personal matter involving the value system of the theorist, our values and social environment influence us ask: What is interesting about the topic? Why is that an important topic? on’t choose a too broad subject working on a abstract level, might be dangerous because of the many details and vagueness think about the population you want to theorize participatory action research: core search by and for those who are to be helped, identify worthwhile problems on which to work is potentially useful Literature reviews a comprehensive literature review serves as a useful source of ideas, and is an essential prerequisite for scientific research. may narrow your thinking rovides much-needed focus and clarity before embarking on the theory construction process Heuristics for generating ideas: help you think about a phenomenon in ways that are different from how you might otherwise proceed. emergent theories: theorist should set aside their preconceptions when studying a phenomenon and let ideas about the concepts and the relationship between concepts emerge as they embed themselves in the surroundings and context of the groups under study. Stay close to the data. Heuristics:

Analyze your own experiences: reflect on what factors have influenced the outcome variable or the phenomenon you are trying to explain based on them Use case studies: in detail and generate ideas based on this study Collect practitioners rules of thumb: research experts who are dealing with the problem (mental) Role playing: put yourself in the place of another and anticipate how this person might think or behave with respect to the outcome variable Conduct a thought experiment: hypothetical experiments or studies that are conducted in the mind as if you have collected data and then imagine the results.

Counterfactuals: focus on what might have been questions. –> can sensitize theoretical possibilities we might otherwise have ignored. Try to frame the phenomena with which you are working. Engage in participant observations: observe others in the situation while you actively participate in those situations. Analyze paradoxical incidents: isolate and analyze, can be pursued mentally by using role playing, analyzing own experiences, thought experiences Engage in imaging: visualize relevant behavior and situations Use analogies and metaphors: apply logic of another problem area to the logic of the area of interest or drawing upon a metaphor.

Imagistic simulations and visualizations often used. generate analogy from well-known principle, recognize it is an example from a well established- principle modifying the original problem situation thereby changing features of it that where assumed to be fixed through association in memory, whereby reminded of, or recall analogous case in memory establish the validity of the analogy in relation to the original problem seek to understand the analogous case apply findings to the original problem

Reframe the problem in terms of the opposite: reversing the focus of your thought to a focal opposite Apply deviant case analysis: explain why certain groups or individuals stand out Change the scale Focus on processes, sets of activities that unfold over time to produce change, maintain equilibrium in a system, or to get from A to event B, or focus on variables, thinking in dynamic processes. Try the opposite of what you normally do. Change nouns into verbs.

Stop the clock, freeze the process at a given point and describe the system in detail at the frozen moment. Process orientated processes, Turner: separation: person becomes detached from a fixed view point in the social structure marginality: person is in ambiguous state, between old and new state aggregation: person enters a new stable state with its own rights and obligations Consider abstraction or specific instances: think about different levels of abstraction. Seek commonalities among phenomena and be skeptical about distinctions.

Different phenomena are often driven by the same underlying principle. Generalize across domains. Ask yourself what is gained and what is lost by moving across different levels of abstraction? Make the opposite assumption and try to gain support for them Apply the continual why and (what do you mean with that? ): ask over and over again Consult your grandmother and prove her wrong, take the obvious and think how it could be wrong. Extending the obvious to the non-obvious.

Push an established finding to the extremes: too much of anything eventually backfires or starts to produce opposite effects Read biographies and literature, and be a well- rounded consumer: can be a rich source of ideas about many facets of human behavior Identify remote and shared/differentiating associates: try to identify as many causes and consequences of it as you can. Free association will include remote associates, things that are unlikely for people to think but might be inspirational. Think of similarities and difference between situations where the phenomenon occurs and when not.

Shift the unit of analysis: focus on units of analysis or on individuals. Shift the temporal dimension. four relational models: female influence, male influence, shared influence, interactional influence model. Shift the level of analysis: proximal (more immediate determinants of behavior) vs distal (influence behavior through immediate determinants) Use both explanations rather than one or the other: consider the fact that al explanations are operating instead of a recency- effect, first is frame, or discount of information. ariation: consider the fact that something can be influenced by more processes, dual- process theory/ framework Capitalize on methodological and technological innovations: Focus on your emotions What pushes your intellectual hot button? Is it worth pursuing Seven stages when processing information (Wyer) attending information when it is encountered interpreting and organizing the information relative to preexisting concepts that are stored in memory construing the implications of the information for already acquired knowledge about relevant people and events storing information in memory etrieve information integrating the retrieved information to make a subjective judgement translating this judgement into an overt response Compartmentalize processes to make mini- theories before constructing a broader theory. It is important to search for alternative perspectives. Reductionism: attempt by scientists to identify, break apart, or reduce nature into its natural constituents. Identifying the core can reveal new insights. Expressing a theory uses six different modalities (McGuire): verbal, abstract symbolic, pictorial, tabular, descriptive statistical, inferential statistical. -> expressing the theory through all modalities might help to grasp it better. Becker heuristics: next-step heuristics: what events lead up to your primary event and what are your next steps. building a machine to maintain the status quo: what should you build to maintain the current level where, when or how question doubt everything someone in a position with authority says, accept nothing for a given 13 core thinking tools that creative geniuses use as they approach the process of idea generation (pp. 71)

Instantiation: a deliberate process that involves specifying concrete instances of abstract concepts in order to help clarify their meaning. Hypothesis: a statement that is derived from a theoretical expression, is more concrete than the originating theoretical expression, and is tied to the empirical realm in such a way as to permit a test of the originating expression. Shared meaning surplus meaning Ways to construct conceptual definitions: search for previous made definitions use the dictionary list key properties ask what do you mean by that? write the definition is you were writing an article explain to others

Multidimensional constructs: more than one definition, subtypes/ sub dimensions Sometimes social scientist create variables for the purpose of building new theories translate an individual- level variable into a contextual- level variable. reframe environmental or contextual variables to represent perceptions on the part of the individual, how does someone perceive the environment? Operational definitions: central to the design of empirical tests of a theory. Consider conceptual definition, than think about the kind of operations and procedures that might be employed to provide a satisfactory indication of the concept in question.

More concrete. Operationism: abandon conceptual definitions, stick to operational definitions, minute changes in methods would produce new concept, and there is no single construct. Strict adherence to operationalism means that both the result of an and the conclusion the investigators derived from it can never transcend the methodology employed. Thus generalization is abridged. Steps when using a thought experiment to clarify the presumed relationship between two or more categorical variables create contingency table, rows contains causes, columns contain effects think of 100 hypothetical people pecify how many you think will fall into each category Relationship between two quantitative variables use a hypothetical scatterplot as a thought experiment. Non linear relationships: inverted U function, (inverse) S shaped relationship Scientist typically think in terms of linear relationships because statistical tools focus on linear relationships general preference for parsimony, it is easier to see the bigger trend (instead of s shape) population partitioning, population chosen is linear When cause is categorial and effect quantitative: hypothetical means/ contingency table, cause listed in rows and scores in the column.

Stating trends in the assigned numbers across the different row categories. When cause is quantitative and effect categorial: shorthand hypothetical scatterplot/ probabilities: turn categorial variable into a quantitative representation so that a scatterplot can be drawn. X-axis is the cause, Y- axis is the probability Moderate relationships: involve three variables and focus on cases where the strength or nature of a relationship between two variables changes depending on the value of a third variable Use hypothetical factorial design: identify the cause, the effect, and the moderator variable.

Than construct a factorial table (2×2 table) in which the levels of the moderator value are listed as columns and the levels of the focal independent variables as rows. Fill it in. Last, calculate the effect of the focal independent variable at each level of the moderator variable at another level of the moderator variable and than calculate the interaction contrast to determine if there is a moderated relationship. Hypothetical factorial design with quantitative variables: group the values of the quantitative variable into high, medium, and low categories.

Hypothetical scatterplots and quantitative variables: if each variable of the moderate relationship is quantitative, focus on describing the differences in the slopes of the two scatterplots. main effect in the factorial table of hypothetical means focus on the marginal means for a given variable, can see what kind of relationship there is, inverse, (non) linear Simple main effects: assign one of the variables the role of the focal independent variable and the other the role of the moderator variable.

Refers to the effect of the focal independent variable at a given level of the moderator variable. Interaction contrasts: compare the effects of the focal independent variable at one level of the moderator variable with the effects of that focal independent variable at another level of the moderator variable. Sometimes might be able to generate reasonable logic for several patterns of entries. Moderator is the dependent variable