Assessment for "counseling" can have different implications depending upon the setting and objectives involved. For example, counseling can range from the provision of clinical services to guidance in selecting courses.
In the case of delivering clinical services, the logical stages are identification of the problem(s), which may include determining diagnoses. This would be followed by determining what the individual's acceptable goals are in developing a treatment plan guided by the assessments required for such a plan.
An example in addressing substance use disorders would be to determine the type and severity of the substance use problems and any co-occurring conditions in accordance with the DSM-5 or ICD-10 criteria. Having identified a diagnosable condition or conditions, the six dimensions of the ASAM Criteria for treatment planning would guide the treatment plan.
Regardless of theoretical orientation counselors need ultimately to select or develop measurement procedures in order to operationalize key constructs to investigate. In Couple and Family Counseling, family assessment literature indicates a great diversity of instruments, but in reality family relationships maybe
most clearly and parsimoniously differentiated along by a three-factors structure of family dimensions for parallel assessment of marital and parent-child relations: marital interaction, mother-child interaction, and father-child interaction ( Jacob & Windle, 1999). The unit of analysis and the assessment context should not be confused. As unit of analysis we can study an individual, a dyad or a whole family but the assessment context is determined by the collection of individuals present for the assessment. Several studies have showed that data drawn from different contexts produce different results but few studies have used observational methods, most rely on self-report measures from a single source. Family-system measures are not a global description of the whole family but a set of measures that include individual, dyadic, and triadic assessment of family observed during whole-family interactions.
Although it is necessary to use self-report measures to tap phenomenological constructs of importance, to anchor these reliable measures to the most reliable and valid data we need the observation of interactions. For example, there is a proliferation of non-standard measures of marital quality and satisfaction, conceptualized simply as a judgement made by spouses at one point in time rather than a trajectory that reflects fluctuations in marital evolution over time (Bradbury, Finchman & Beach, 2000).
All of the above are great answers. I would recommend after identifying a specific problem (e.g., depression, physical problems), determine incidence (when did it first occur), recency (when did it last occur), breadth/severity (how many symptoms/how bad is or was it), and service utilization (including when, how long and type of service provided). Also, when using a semi-structured instrument, don't fill in answers for the.client. For example, if the client says "a couple", ask "Do you want me to write down 2, 3 or some other number?" so the information is accurate.