Professional Documents
Culture Documents
Last Topics Developmental Design Application Analysis
Last Topics Developmental Design Application Analysis
in behavior. They aim to describe the relationship between age and other variables, such as language
ability. There are two main types: cross-sectional design and longitudinal design, each with its strengths
and weaknesses. For instance, studying how language ability changes with age suits such research.
using separate groups of participants. For example, a researcher wanting to look into the connection
between IQ and aging could choose three different age groups - 40s, 60s, and 80s - and measure IQ for
each one (see Figure 10.7). It allows researchers to observe behavior changes as people age without
waiting for participants to grow older, collect data quickly, and require no long-term cooperation. The
cross-sectional research design has flaws, such as the inability to predict individual development over
time and generation effects, which may distinguish groups based on factors other than age. Internal
validity may be jeopardized due to the apparent relationship between age and other variables caused by
generational differences.
The longitudinal developmental research design involves measuring a variable in the same group
of individuals over time, typically every few months or years. These individuals are usually cohorts of
roughly the same age and have grown up in similar circumstances. Multiple measurements of a particular
variable are made at two or more times to investigate the relationship between age and that variable. For
example, to examine IQ and age using the longitudinal approach, a researcher might measure IQ in a
group of 40-year-olds and then measure the same individuals again at ages 60 and 80 (Figure 10.8). The
longitudinal developmental design is a research method that examines one group of people over time,
avoiding cohort effects and examining behavior changes with age. However, it is time-consuming,
expensive, and subject to high dropout rates, known as participant attrition or mortality. This can weaken
the internal validity of the research, as participants may lose interest, move away, or die. The group at the
end may have different characteristics from the group at the beginning, as motivation may explain
changes over time. Additionally, the same individuals are measured repeatedly, potentially affecting
scores obtained later in the study due to previous experience with the test or measurement procedure.
This chapter discusses the application and analysis of between-subjects and within-subjects
designs, with the exception of the quasi-experimental pre-test–post-test non-equivalent control group
design. Two-group designs offer simplicity and an easy understanding of results but do not provide the
full functional relationship between variables. Statistical analysis for numerical data involves comparing
means with a t-test or single-factor analysis of variance, while chi-square tests are appropriate for non-
numerical data. The pre-test–post-test non-equivalent control group design is used for numerical data
analysis. A two-factor, mixed design analysis of variance is appropriate, which is available on most
statistical software programs like SPSS. If comparing pre-post means for one group, a repeated-measures
t-test can be used, while an independent-measures t-test is appropriate for comparing the two group
and measuring a dependent variable (scores) in each condition. No independent variable is manipulated in
non- and quasi-experimental research, but groups of scores are compared. Non-equivalent group studies
compare scores between groups, whereas pre-post studies compare scores before and after treatment. The
variable that distinguishes groups is commonly referred to as an independent variable, but it is more
score obtained by each participant. In non-equivalent control group studies, the quasi-independent
variable is determined by the difference between the treatment and non-treatment groups. In time-series
studies, the quasi-independent variable is 'before versus after treatment'. This concept is used in both non-
experimental and quasi-experimental research. In differential research, the participant variable is used to
differentiate groups. For example, in a study comparing self-esteem scores between two-parent and
single-parent homes, the number of parents is the quasi-independent variable, while self-esteem is the
dependent variable. In a developmental study, the different ages are the quasi-independent variable, and