Term Paper Instructions

1. Name your independent and dependent variables. State the numeric values they take on and describe precisely how they are scored on the California Family Risk Assessment. State the maximum and minimum possible numeric scores a household can be given on both the independent and dependent variables and state how a household gets both the minimum and maximum score on each variable.  To be able to do this, examine your item on the CFRA. Describe how the probability of maltreatment changes or doesn’t change as a function of risk scores.
Hints: For all independent variables on the California Family Risk Assessment (CFRA) and for the dependent variable you all share, the lowest possible numeric value or “score” is zero (0).  You can see this for independent variables by looking at the CFRA form on BB.  Every lettered response alternative a., b., etc., has a number that is its numeric value shown to its right on the form. “Response alternatives” are the same as “variable values” or “values of a variable”.
More hints: The following independent variables on the CFRA all have two possible response alternatives (i.e. variable values), lettered a. and b. and have maximum possible numeric values or scores of one (1):  N1, N3-N7, A1 – A4, A8, and A9.  A6 also has two possible response alternatives (i.e. variable values) but has a maximum numeric value or score of two (2). “Numeric values” are the same as “scores” or “risk scores”.
More hints: One CFRA independent variable, N2, has a maximum possible score of three (3) and the person calculating the score for N2 is instructed to “assign the highest score that applies”. (all the cases (households) in the databases you have used for your analysis were scored by child welfare workers on real cases in the field—i.e. out there in reality land).
More hints: Six CFRA independent variables have maximum possible scores arrived at by following an instruction like this:  “Check applicable items and add for score”.  To see this, look at the CFRA form. These variables are N8 – N10; A5, A7, and A10.  The maximum possible scores for these variables vary from two (2) (for variable N8) to three (3) (for variables N9, N10, A5, A7, and A10). All these variables get their maximum possible scores when all “applicable items” are checked or noted by the scorer as present in the case.  These response alternatives b., c., etc., are distinct from response alternative a., which, for these variables is always designated “not applicable”.
More hints: The dependent variable has a minimum scored zero (0) for households that have no substantiated maltreatment of their children during a two-year follow-up period subsequent to assessment with the California Family Risk Assessment.  The dependent variable has a maximum score of one (1) for households that do have substantiated maltreatment of their children during a two-year follow-up period subsequent to assessment with the California Family Risk Assessment.
Final hint:  For independent variables with only two response alternatives, a. and b.,  the minimum possible numeric score is always zero (0).  For these variables the maximum possible score is the numeric score given for response alternative b.  In all cases this will be either one (1) or two (2).  If your independent variable has only two response alternatives, look at  the CFRA to see if its maximum score is 1 or 2. For these CFRA variables, a household “gets” a score of 0 if it is accurately described by response alternative a. and it gets a score of 1 or 2 if it is accurately described by response alternative b.  When you describe how your 2-reponse alternative independent variable (if you have one) gets scores of 0, 1, or 2, you must state the exact response alternative associated with the minimum and maximum possible scores on your variable.  Example for N3 on the CFRA: “Households get a minimum score of zero (0) on my independent variable if they have not previously had CPS services (voluntary or court-ordered). Households get a maximum score of one (1) on my independent variable if they have previously had CPS services (voluntary or court-ordered).
Example statement of minimum and maximum possible scores for a CFRA variable and how a household “gets” these scores:
My independent variable is N9, named “Characteristics of Children in the Household”.  This variable takes on the following values:
a. Not Applicable (numbered/scored 0); b. Medically fragile/failure to thrive (numbered/scored 1); c. Developmental or physical Disability (numbered/scored 1); d. Positive toxicology screen at birth (numbered/scored 1).  The variable is given a risk score for a household by following the instruction printed with the variable to “check applicable items and add for score”.  To do this, the person scoring the household would add up the numeric values for each child characteristic present among the children in the household being scored.  A household in which all three of the conditions/characteristics indicated by response alternatives b. through d. were present would have the maximum possible score = 1 + 1 + 1 = 3.  The minimum possible risk score for my independent variable is 0, for households described by response alternative “a. not applicable” whose children have none of the conditions described by response alternatives b. through d.
2. State your hypothesis about the relationship between these variables:
Example:  My hypothesis is that groups of households that have higher risk scores on my independent variable—i.e. households in which more of the conditions described by alternatives b. through d. are present–will be exhibit higher percentages (probabilities) of experiencing maltreatment during follow-up. Your hypothesis will have the same form, i.e. you should hypothesize that higher numeric scores on your independent variable are associated with higher percentages (probabilities) of experiencing maltreatment during follow-up.
3. Paste both your SPSS output containing your crosstabulation of your independent and dependent variables and your SPSS output containing your Chi Square statistic into your paper following #2 (your statement of your hypothesis).
4. Paste your Excel chart graphically depicting the relationship between your independent and dependent variable following #3.
5.  Write a paragraph interpreting your analytic findings.  State whether the actual findings of your analysis support, or fail to support your hypothesis.  State whether or not the direction of the relationship shown in your Excel chart is what you expected it to be, or not, according to your hypothesis.  State whether your Chi Square output indicates that the relationship between your independent and dependent variables is statistically significant or not.  Statistical significance is the bottom line.  If the direction of the relationship is what you expect according to your Excel chart but the relationship is not statistically significant, you will state that the findings are INCONSISTENT with your hypothesis.
REQUIRED COMPUTER OUTPUT, EXCEL CHART, AND EXAMPLE INTERPRETATION APPEAR BELOW. YOUR INTERPRETATION MUST FIT YOUR FINDINGS, NOT THE FINDINGS INTERPRETED BELOW.

recurrence CD–Follow-up Subatantiation Within 2 Years * n_nine CD–N9–Characteristics of children in the household–see response alternatives Crosstabulation    n_nine CD–N9–Characteristics of children in the household–see response alternatives    Total
.00 not applicable    1.00 one of: med frag, dev-phys-disability, pos tox    2.00 two of: med frag, dev-phys-disability, pos tox    3.00 all of: med frag, dev-phys-disability, pos tox
recurrence CD–Follow-up Subatantiation Within 2 Years    1.000 Substantiated maltreatment within 24 mos. of index investigation    Count    370    152    27    5    554
% within n_nine CD–N9–Characteristics of children in the household–see response alternatives    19.6%    28.3%    36.0%    62.5%    22.1%
.000 No subsequent substantiated maltreatment    Count    1520    386    48    3    1957
% within n_nine CD–N9–Characteristics of children in the household–see response alternatives    80.4%    71.7%    64.0%    37.5%    77.9%
Total    Count    1890    538    75    8    2511
% within n_nine CD–N9–Characteristics of children in the household–see response alternatives    100.0%    100.0%    100.0%    100.0%    100.0%

Chi-Square Tests    Value    df    Asymp. Sig. (2-sided)
Pearson Chi-Square    34.862a    3    .000
Likelihood Ratio    31.830    3    .000
Linear-by-Linear Association    33.534    1    .000
N of Valid Cases    2511
a. 1 cells (12.5%) have expected count less than 5. The minimum expected count is 1.77.
OUTPUT CONTINUES ON FOLLOWING PAGES

Example Interpretation of findings:
Groups of families whose children exhibit HIGHER NUMERIC SCORES AND, THUS, MORE OF THE CHARACTERISTICS DESCRIBED BY RESPONSE ALTERNATIVES B. THROUGH D. EXHIBIT HIGHER RATES OF SUBSTANTIATED MALTREATMENT DURING FOLLOW-UP.  This relationship is statistically significant with a Chi Square value of 34.86 and associated Asymp. Sig. value of .000.  The results are consistent with my hypothesis, as stated above.