Introduction

Iowa Vocational Rehabiliation Services provides resources and training to individuals with disabilities in order to facilitate better job placement. The dataset analyzed contains information on cases where services were terminated, either due to successful completion of the program, or individual opt-out.

The general goal of this analysis is to understand sucessful and unsuccessful outcomes related to the program. After a general overview is achieved, individual analysis focusing on demogramics, disability types, and geographic locations will be explored.

Success Definition

Outcomes can be treated as successful when an individual increases their hourly income, and unsuccessful when individuals opt-out of the program or see no pay increase. The dollar increase in income can be used to measure level of success. Specifically:

  1. In general, success is considered as any case where an invidual increased their hourly pay at the time their account was closed.

  2. Recieving an hourly pay raise greater than 11 dollars is considered high level success. This corresponds to the last two categories of success level.

Data Exploration and Overview

Program Level

Generally, the program has a higher number of success than failures (Figure 1). However, of these successes, there are a higher number of low level pay increases (between 0 and 10 dollars), than there are high level pay increases (greater than 11 dollars) (Figure 2). Overall, the number of instances in each category decreases as the pay increases, however, this is distored in the overall display due to the fact that those without a job will start at minimum wage. Figure 3 demonstrates this effect.

Suprisingly, there are cases where individuals without jobs do start below minimum wage, however, this could be due to employment in service industries which supplement hourly wages by allowing for tipping. There are also individuals that start the program with jobs and are still able to increase their hourly pay by 15 dollars or more (Figure 3).

## `summarise()` has grouped output by 'success_level'. You can override using the `.groups` argument.

Number of Opened Cases

In general, there appears to be a negative relationship between the number of cases opened and the proportion of successes. This can be seen by the fact that the number of opened cases has been increasing since 2012 (Figure 4), while the proportion of successes has been decreasing (Figure 5). Also, in almost every year for which the number of opened cases exceeded the average (Figure 4), the proportion of successes was less that 50% (Figure 5).

Future analysis may explore if this relationship is due to the added pressure on staffing when case load increases. This would require data to be gathered on the amount of time allocated for clients, which is currently unreported.

## `summarise()` has grouped output by 'fiscal_year', 'gender'. You can override using the `.groups` argument.

Demographic Exploraiton

The number of males and females involved with the service, as well as the proporiton of success for each, is relatively similar (Figure 6). There does appear to be some slight variation in the proportion of high level outcomes, with males seeing a greater proportion of 11 dollar or greater hourly pay increases.

The number of individuals recieving services who are listed with minority status is far less than the number with non-minority status (Figure 7). Those with minority-status also have a higher proportion of failures, and a smaller proportion of high level outcomes, compared to those with non-minority status.

## `summarise()` has grouped output by 'gender'. You can override using the `.groups` argument.
## `summarise()` has grouped output by 'minority_group'. You can override using the `.groups` argument.

Disability Type

While the Cognitive Impariment category has the highest proportion of success (Figure 8), they have the lowest proportion of high level outcomes. This group also has a relatively low number of individuals entering the program with a job (Figure 9), meaning that their success could be largely due to the minimum pay increase associated with employment.

The Visual and Communicative Impariment category has the second highest level of success, with the highest proportion of high level pay increases (Figure 8). However, as they also have a relative low number of individuals entering the program with a job (Figure 9), some of this may be attributed to the inital increase in wage from simply being employed.

The Physical Impariment category has the highest proportion of successes recieving over 15 dollar pay increases (Figure 9), however, also have one of the lowest overall proporiton of success (Figure 8). This indicates a need to further explore the trends within this category, possibly identifying the distinctions between the high level and failure subgroups. If such knowledge was gained, it might more efficiently inform the intervention strategies relevant to each individual subgroup.

The Deaf and Hard of Hearing categoy seems to have the best overall outcomes, as they have a relatively high proportion of successes (Figure 8), and also have a high proportion of high level success. This is despite the fact that many of them begin the program with a job (Figure 9).

## `summarise()` has grouped output by 'disability_type'. You can override using the `.groups` argument.
## `summarise()` has grouped output by 'disability_type'. You can override using the `.groups` argument.
## `summarise()` has grouped output by 'disability_type'. You can override using the `.groups` argument.

Duration

On average, most disability categories have case durations within 30-40 months (Figure 10). The category which seemed to have the best overall success, Deaf and Hard of Hearing, has the lowest average duration, while the Cognitive Impariment category has the highest. This does indicate the potential for a negative relationship between duration and high level success.

Interestingly, the range of duration is between 0 to over 250 months, while the IQR for almost every category is around 25 months. This might indicate a need to futher explore the characteristics of individuals within the data to determine if there are relationships between the outliers, or distinctions between potential subgroups.

For most disability types, wage change tends to increase with duration until approximately 150 months (Figure 11). After this point, wage change and time are negatively related. The exception to this trend is the Cognitive Imparments disability category, whose trendline demonstates a dip around 150 months, but continues to rise after that point.

This likley indicates a need to further explore the relationship between duration and outcome, potentially identifying the length of time which generates the highest probability for success by disability type. This might allow staff to more efficiently prioritize their time to address those cases which are reaching that threshold.

Location

Office Area

In general, the offices have a farily similar distribtuion for most disability types, however, they vary substantially in the number of cases they accept (Figure 12). There is a slight variation in the proprtion of successes between offices (Figure 13), as well as the average duration of their associated cases (Figure 14).

The Central Office has the highest level of success (Figure 13), as well as one of the lowest average montly case duration (Figure 14). Ottumwa has the lowest proportion of successes (Figure 13), however, has a relatively usual average case duration (Figure 14). Ottumwa's distribtuion of disability types may be effecting their sucess proprtion, as they have a low number of cases in the Cognitive Impariment category (Figure 12), which has the highest proportion of successful instances.

## `summarise()` has grouped output by 'office_area'. You can override using the `.groups` argument.
## `summarise()` has grouped output by 'office_area'. You can override using the `.groups` argument.

Client County

Generally, the Northern Counties in Iowa have a higher probability of sucess than other areas, and counties in the South Eastern area of the state have the lowest (Figure 15). These probabilites do not appear to be related to the population of the county, as there does not appear to be any obvious trends within Figure 16.

There does appears to be some spatial effect outside of the county level, as there are clusters of counties that exhibit similar probabilites of success. This could be due to the fact that people in close geographic proximitity are more likley to be supported by the same office. This may indicate that the office an individual attends may be more important to success than the geographical location of the participant themselves. Further analysis may focus on exploring this hypothesis.

Conclusion

Generally, gender, minority status, disability type, and office are related individual's outcomes within the rehabilitation program. While gender only plays a relatively small role, the other variables seem to have a strong relationship with both the probability of success as well as success level.

Males, individuals with non-minority stats, and individuals within the Visual and Communicative Impariment category have the highest proprtion of high level successes. This indicates a need for the program to look into the differences in services provided to these invdividuals, comparing them to those of other types, as well as look into the general biases associated with their employment targets.

While those within the Cognitive Impariment category have high proportions of success, they also have extremly low proportions of high level success. While this may indicate a need to further explore potential employment opportunities, it must also be analyzed from the perspective of the goals of the individual. A better method for such analysis would be to explore the level of satisfaction the individuals have with their final employment placements.

With regards to the agencies, there appears to be a negative relationship between the amount of cases opened and the proprtion of successful closed cases. This indicates a need to further explore the funding and staffing of these agencies, possibly using the Central Office, which had the highest level of success, as a baseline. Additional variables, such as time spent with advocates, would help to further understand this relationship.