Question

The data are not actual data, but are modeled after data collected by Trautwein and Ammerman (2013) for a study on the effects of cochlear implants (CI) on the language, speech, and perception skills of 32 deaf children: 16 subjects with at least one cochlear implant (CI) and 16 subjects (with hearing loss) without an implant. (CI subjects were implanted at 13 or 14 months of age. So, no subjects had CIs at the time of the first data collection.) Auditory perception and speech production skills were measured at several points over a three-year period. Here, I include the data collected at 1 year of age and 3 years of age. The PLS was used to measure language before the study, and the CELF-P was used in subsequent years. The data set code book is provided after the questions.

The data set contains information of demographic variables and assessment variables. Here I briefly describe the assessments. If you wish to know more about the assessments, you can easily find additional information on the internet.

The Early Speech Perception (ESP) assessment contains single-word targets that differ in number of syllables and stress. This tool uses a series of subtests to assess pattern and word discrimination in a closed-set format. The results of this test allow the placement of children into four speech perception categories: 1=no pattern perception, 2=pattern perception, 3=some word identification, and 4=consistent word identification.

The Preschool Language Scale-5 LS-5 measures comprehensive language in children from birth through age 7.11. Scores include Auditory Comprehension (AC), Expressive Communication (EC), and a Total Language (TL) score. All scores have a mean of 100 with a standard deviation of 15. The TL score is typically the most representative of a child’s overall language functioning. The AC score probes aspects of comprehension whereas the EC score measures oral language expression.

The Clinical Evaluation of Language Fundamentals, Preschool Edition (CELF-P) evaluates a wide range of language skills in children aged 3.0 to 6.11. Scores include the Core Language Score (CLS), Receptive Language Index (RLI), Expressive Language Index (ELI), Language Content Index (LCI), and Language Structure Index (LSI). All scores have a mean of 100 with a standard deviation of 15. The CLS is typically the most representative of a child’s overall language functioning. The RLI probes aspects of comprehension whereas the ELI measures oral language expression.

Finally, all participants were assessed using the Goldman Fristoe Test of Articulation at the conclusion of the study.

The Goldman Fristoe has a mean of 100 and a standard deviation of 15.

1. Trautwein and Ammerman were curious to see if they could obtain permission to gain information on the SES of the families whose children participated in the study. They theorize that SES, communication modality, and cochlear implant status, separately and together, influence language outcomes.

What type of inferential test should be used to test the hypothesis? Why?

Name the independent and dependent variables.

Write the research questions that the researchers wish to answer.

If all null hypotheses were rejected, what would the researchers conclude?

How will they know which combination of factors has the greatest influence on language outcomes?

DEHS Data Set for Assessment Projects
INTD 5064 Applied Statistics for Health Care Practitioners
CI Subjects
Subject ID	Group	Gender	Mode of Communication	ESP-Pre	TL-Pre	AC-Pre	EC-Pre		ESP Post	CLS-Post	RLI-Post	ELI-Post	LCI-Post	LSI-Post	GF-Post
C01	1	1	1	1	86	89	95		3	95	96	99	90	88	101
C02	1	2	1	1	82	85	85		3	85	89	87	83	78	89
C03	1	1	1	1	85	78	80		3	90	92	88	88	84	85
C04	1	1	1	2	81	81	82		2	80	88	84	79	76	105
C05	1	2	1	1	92	95	100		4	100	100	95	98	104	82
C06	1	1	1	1	79	80	72		4	92	94	90	92	82	90
C07	1	1	1	1	78	80	74		3	84	88	79	86	80	80
C08	1	2	1	1	111	115	102		3	120	122	112	109	110	102
C09	1	1	1	2	87	89	79		4	88	92	87	87	82	82
C10	1	2	1	2	99	98	89		4	100	98	90	104	106	95
C11	1	1	1	1	67	69	68		3	79	84	82	70	78	77
C12	1	2	2	1	102	100	98		2	95	98	93	96	90	82
C13	1	1	2	1	74	74	70		2	70	72	70	71	68	74
C14	1	2	2	1	68	65	70		1	67	69	66	65	68	70
C15	1	1	2	1	95	90	101		2	90	92	88	85	95	87
C16	1	1	2	1	75	77	76		2	72	75	70	74	68	69
Non CI Subjects
Subject ID	Group	Gender	Mode of Communication	ESP-Pre	TL-Pre	AC-Pre	EC-Pre		ESP Post	CLS-Post	RLI-Post	ELI-Post	LCI-Post	LSI-Post	GF-Post
S01	2	2	2	1	95	89	100		2	93	89	85	100	92	82
S02	2	1	2	1	87	85	95		1	80	83	82	80	78	70
S03	2	2	2	2	76	78	80		1	78	72	70	77	80	65
S04	2	1	2	1	75	73	72		2	73	71	72	75	72	68
S05	2	1	2	1	79	78	80		1	80	70	72	85	81	63
S06	2	1	2	1	77	78	74		1	75	77	68	85	70	71
S07	2	1	2	1	78	80	72		2	72	75	70	72	69	60
S08	2	2	2	1	77	81	75		1	80	89	73	75	78	65
S09	2	1	2	1	87	88	84		1	80	84	80	75	80	68
S10	2	1	2	1	89	98	82		3	90	100	85	85	85	78
S11	2	2	2	1	90	94	92		1	85	90	82	79	87	72
S12	2	1	2	1	73	78	72		1	70	74	67	79	68	66
S13	2	2	2	1	74	74	75		1	70	72	69	67	75	67
S14	2	2	2	1	68	65	68		2	62	60	61	67	60	55
S15	2	1	2	1	75	74	72		1	71	77	67	75	71	70
S16	2	1	2	1	74	77	73		1	77	74	68	80	72	65

Answer 1

1.      For each variable listed below, give the scale of measurement and the reason for your answer.

a.      Group - Nominal because group does not have any order. We cannot say if group 1 is greater than group 2.

b.      Gender - Nominal

c.      Age - Ratio scale. It has order and age zero means actually zero

d.      ESP- Pre - Interval scale

e.      CLS-Post

2) Mean and Standard deviation cannot be measured for nominal and ordinal scale variables. So we cannot calculate standard deviation for gender, group, mode of communication

3) The Goldman Fristoe is same across all groups, gender and mode of communication.

We can use ANOVA to test the hypothesis