General psych: want to describe and explain behavior

Developmental psych: want to describe and explain behavioral change over time

Example: skepticism
General psych: observe or survey skepticism in college sophomores.
Describe average amount of skepticism
Look for factors associated with skepticism --
IQ, SES, gender, childrearing, etc.

Developmental: Does average amount of skepticism change with age?
If so, what explains this change?
Goal is to describe developmental function (pattern of age change)

 
 
 
 
 

Picture suggests people generally more skeptical as they age
& this line describes normative (average) course of change

Also suggests young persons will be as skeptical as older when they become old

BUT, not easy to define developmental functions: method problems

Three dimensions in developmental research:
Age(s) of participants
Time(s) of measurement (historical time frame in which study is conducted)
Cohort(s) of participants: (historical time frame of birth)

Developmental psy. traditionally concerned with changes occurring in relation to age.
Hard to separate from time and cohort dimensions:

Traditional designs: cross-sectional & longitudinal

Cross-sectional design:
One time of measurement
Two or more ages
Two or more cohorts
Age and cohort overlap completely

 
 
 
 
 
 
 
 
 
 
 
 

Does this give us normative age change?
No: have cohort problem -- can’t say those born in 1981 will behave the same way as
those born in 1941 when the 1981 cohort is age 60. Maybe change in childrearing
philosophy discouraged skepticism more in younger cohorts.
 

Longitudinal design
Two or more ages
Two or more times of measurement
One cohort
Age and time of measurement overlap completely

 
 
 
 
 
 
 
 
 
 
 
 

Does this tell us how skepticism changes normatively with age?
No. Can’t generalize from this particular cohort to persons born earlier or later.
Childrearing philosophy could change again & encourage skepticism for later cohorts
 

Sequential designs
Two or more ages
Two or more cohorts
Two or more times of measurement

 
 
 
 
 
 
 
 
 
 
 
 

Examples of traditional and sequential designs within graph:

Traditional cross-sectional: Study ages 20, 40, & 60 in 2001
Cross-sectional sequence: Study ages 20, 40 & 60 in 2001 and 2021
                                          (different subjects in 2021 than in 2001)

Traditional longitudinal: Study the 1981 cohort in 2001, 2021, and 2041
Longitudinal sequence: Study the 1981 cohort from 2001 to 2041
    and the 2001 cohort from 2021 to 2061 (same subjects each time)

Most efficient:
Use both cross-sectional and longitudinal sequences.
Compare longitudinal and cross-sectional findings.
If findings are similar, closest one can come to age-related change
If findings differ, look for cohort differences/ other factors that might explain diff.

Example: Cross-sectional, longitudinal findings on age differences in adult intelligence differed.
Why? Differences across cohorts in education level. Oldest cohorts had less education.

Most data on aging is cross-sectional. Need to be concerned about cohort differences.

Advantages and disadvantages to each type of design.
 

	Cross-sectional	Longitudinal	Sequential
Advantages	Inexpensive Easier to do Quicker Good as first step OK if only want to know current age/ cohort differences (ex. how people of different ages will vote in 2002)	Can assess intra- individual change (within a single cohort of persons over time) OK if not concerned about time of measurement differences  (ex. short term study of change in infant visual acuity)	More precise Can generalize across time and cohorts Can study time and cohort effects in  addition to age  effects
Disadvantages	Can’t separate age effects from cohort effects. Apparent differences related to age may be cohort differences.	Expensive Difficult to do Time consuming Selective sampling Selective drop-out of    participants Retest effects Can’t separate age   & time effects (can’t    generalize to earlier   or later born cohorts   who live through   different times or    through same times   at diff. ages)	Expensive Difficult to do Time consuming

Cross-sectional, longitudinal, sequential are general data collection designs.
  Can use any type of method: experiment, survey, interview, observation, etc.

Experiments can help to solve questions about cohort differences.
Ex. if education or practice improves performance of older more than younger
initial difference was probably due to differential experience (a cohort effect)

Cross-sectional designs typically involve comparing mean (average) scores on some measure
    across different age/cohort groups.

Studies of change versus continuity/stability in individuals over time are always longitudinal &
   usually involve correlations of individual scores across time.
   High correlation = normative stability (person ranked similarly in cohort over time)

  Similar score at each time = level stability (persons in cohort (average scores) don’t change in either
     direction)

  Need to assess both normative and level stability in studies of continuity versus change.