Testing is the sample size

Project description
There are two assignments needed in being done, separately don’t change the topics and followed the instructions thanks.
One page
Respond to this posting in one or more of the following ways:

? Ask a probing question.
? Share an insight from having read your colleague’s posting.
? Offer and support an opinion.
? Validate an idea with your own experience.
? Make a suggestion.
? Expand on your colleague’s posting.

Topic: Parent’s child diagnosed with Autism Spectrum Disorder

The test I am designing is intended to measure attachment in parent-child dyad for parents who have a child diagnosed with Autism Spectrum Disorder (ASD). Traditional

scales measure attachment behavior as indicated by neurotypical childhood behavior anticipated when a child is in attachment relationship with their parent. The Autism

Attachment Intensity Measure (AAIM), allow for children to be considered in attachment relationship without some of the expected behaviors’ such as verbal request or

eye contact. Therefore it is very important that this scale then use a specialized subset of the population. The target population is then a parent of a child

currently under the age of two, affected by ASD, where the child is living with them and has been uninterruptedly so for the last 2 years.
Sample Strategy
In order for a probability sample to be used the researcher must be able to identify each individual within the population and there must be a random, independent

selection, of the the individuals who will participate(Kline, 2005). Due to the Health Information Act (HIA) in Canada, the identification of all of individuals under

the age of 14, living in Calgary, Alberta with a diagnosis of ASD is not possible. In order to adjust to this barrier I will be required to use a nonprobability

sample, which is used when one of the criteria for probability sampling is not met. A snowball sample, in which each participant is requested to provide the name of

another individual who meets criteria (Kline, 2005) is not appropriate. I deem this inappropriate after many years as a clinician with this population, and seeing the

sensitivity needs related to the diagnosis. In order to overcome the lack complete identification of individuals within the population, and HIA requirements, I will

use a convenience sample. In this method, those who are most convenient to locate and participate are used (Kline, 2005). As a result, my study will require that I

utilize advertisement for volunteers, in areas where both families with children with disability frequent (i.e support services) and where neurotypical families are

found (i.e. play areas, pediatricians, recreation). This will allow me to draw from both the families who identify with the diagnosis and those who choose to remain

more focused on mainstream activities. This will not, however, remove a barrier to the generalizability of this sub-population in that only those parents who volunteer

will be used. It could be said that then the AAIM does not generalize to the broader ASD population. Only replication of the study over will assist with building the

support for the AAIM.
Sample size
In order to determine the appropriate sample size for a study, the population size must be known or predicted. For my study I can access the general population

statistics provided by Health Canada, which estimates that 1 in every 150 children is effected by Autism Spectrum Disorder (retrieved July 22, 2014

sc.gc.ca/hc-ps/dc-ma/autism-eng.php), while ASD surveillance estimates the actual number to be 1 in 110 children living with Autism Spectrum Disorder (retrieved July

21, 2014 from percentage of the population estimated to be

0-14 is 17% (retrieved July 22, 2014 from http://www.statcan.gc.ca/pub/91-003-x/2007001/4129904-eng.htm), while the population of Calgary, is estimated at 1, 145, 552

(retrieved July 22, 2014 from http://www.statcan.gc.ca/pub/91-003-x/2007001/4129904-eng.htm). Therefore the estimate of the population of children in Calgary, Alberta,

Canada affected by ASD is approximately 17% x 1, 145, 552 = 194, 744, making the final estimation to be 1, 170 children when using surveillance data. From here,

researchers must select their chosen alpha or confidence level. For a population of 1,170 and an alpha of 5% with a confidence level of 95%, I will need to include 289

parents in the survey (retrieved July 22, 2014 from .

rerences

Kline, T. (2005). Psychological testing: A practical approach to design and evaluation. Thousand Oaks, CA: Sage.

One page

Respond to this posting in one or more of the following ways:

? Ask a probing question.
? Share an insight from having read your colleague’s posting.
? Offer and support an opinion.
? Validate an idea with your own experience.
? Make a suggestion.
? Expand on your colleague’s posting.
? Delores West

Topic: Testing is the sample size

One of the most important considerations that I should look for in my project as well as in any testing is the sample size (Kline, 2005, p. 81). When one examines

population of who is going to be tested, making sure that sample size is not too small or too large is extremely important. Making sure that a sample is collected that

is representative of the population is what will ultimately determine if the measuring tool is adequate and a good representation of validity.

Another area of importance related to sampling is what type of sampling should be used. One type of sampling style is random sampling. In random sampling each member

of a group has an equal chance of being selected, which creates little opportunity of exclusion (Kline, 2005, p. 78). To use random sampling each member of the

population or group would need to be identified and then a selection would be made by a random draw possibly using numbers.

Other types of random sampling are stratified random sampling and disproportionate stratified random sampling. In stratified random sampling one does not leave

everything to chance such as one can group or categorize by age or gender for example (Kline, 2005, p. 78). In this sampling style one makes sure that the sample is

similar to the population in certain respects. In disproportionate stratified random sampling, the sample is not proportioned to the size of the stratum. In my

particular project of testing discipline effectiveness, proportioned random sampling would work best so populations are equal between gender and ages. In my project I

will be concerned with a select population of males and females who have had some type of disciplinary problems in the past or who currently has discipline problems.

Thus the population will need to be proportion in some respect.

One other consideration in sampling that will be examined in my project is missing data. Considering that my testing measure will be in a questionnaire type form,

missing data from unanswered questions or skipped questions will be a concern. Considering that the population that I will be testing is a younger population,

nonresponse missing data will more than likely be a concern. The population may overlook, misunderstand, or simply choose not to answer a particular question for

whatever reason. The response given could also easily be an untruthful answer (Kline, 2005, p. 85). One solution is to look over each questionnaire as it is turned in

and make sure all questions have a selection although, this does not fix all of the problems of missing data.

Reference

Kline, T. (2005). Psychological testing: A practical approach to design and evaluation. Thousand

Oaks, CA: Sage.