Zetterberg says a concept has three elements: (1) term (label) which means (2)a certain conception (definition). Term designates (3)a certain object or event (thing to be measured).
In concept formation the definition specifies what properties are included, specifies the logical relationship between the properties, and specifies logic of procedure for collecting, cataloguing, and manipulating experential data (information on aspects of concept properties).
Note discussion of intelligence (term). Definition ability to adapt to one's environment. Properties of intelligence include cognitive skills and motor skills. Measures of verbal skills are extent of vocabulary and of motor skills that are manipulative of spatial shapes of objects against a contextual field. Logical relationship between properties: assumes that intelligence is tied to chronological timetable, maturational development of skills is age-graded, and individuals may be ranked relative to normative standards of performance for age peers.
Note discussion of social stratification (term). Properties of social stratification include status, class, and power and defined as follows: status-honor bestowed or claimed; class-ability to compete successfully for goods and services in market situation; and power-ability of individual or group to exercise its will over and against the communal effort of others. Measures of status typically include ownership of material goods of varying social worth, of class annual income, and of power positions of authority (legal and illegal). Logical relationship among properties assumes hierarchy (or transitive relationship). [Other status hierarchies include sex, race, ethnicity, religion.]
Mutually exclusive sets vs. fuzzy sets.
Strict interpretation of empiricism. You can only talk about what you measure. If you say intelligence is a two factor intelligence test, that is all it is. You can only make statements in which all your concepts have empirical measures. This assumption limits one's ability to develop theories of a general nature.
Loose interpretation of empiricism. Your concept means more than your operational definition. Masters & Johnson study of human subject response about love. Their definition of love was physical aspect of response, other aspects of love recognized but not studied.
If you have different interpretation for same measure you have a problem about the validity of your measures. For instance, social psychologist who measured alienation and anomie used same items in scales.
Concepts may have single or multiple empirical indicators. Multiple measures permit substitutability.
Issues of reliability - repeatability of results (1) across measures - congruence (multiple measures of fertility); (2) across two or more observers - inter-subjectivity or objectivity (interaction coding schemes); (3) on same measure - precision (multiple observations to obtain mean weight); and (4) across time - constancy POMs profile of Mood States.
If you have reliability you can assume validity -measuring what you think are measuring. Look at your text and ascertain definition of content validity, face validity, construct validity and predictive validity.
Building concepts thru induction. Building from observation to category. Based on repeated observation you come with a category. You define a decision rule for what items fit into your set and what the boundaries of the set are. You may decide to follow rule of mutual exclusivity. If you follow this rule it takes only one nonconforming instance to force you to change boundaries. You may decide to adopt a probabilistic model. It takes a sufficient number of cases to force you to change boundaries. The principle of adequacy lets you argue you have observe a sufficient number of cases to feel psychologically comfortable that you have discovered what fits in a category.